Adaptive Environments Laboratory
School of Architecture and Planning
112 Hayes Hall
SUNY /Buffalo
Buffalo, N.Y. 14214-3087
Edward Steinfeld, Arch. D., Project Director
Gary S. Danford, Ph.D., Senior Research Associate
Joel Zingeser, Building Technology, Inc., Consultant
Laurie Baker, Steven Winter Associates, Consultant
This report documents a review of research literature, products, building codes and facility management practices related to automated doors. It also provides recommendations for revising the Americans with Disabilities Act Accessibility Guidelines (ADAAG). These recommendations provide comprehensive attention to the accessibility needs of people with disabilities. The term automated doors includes full powered automated doors such as those that are commonly used at supermarkets and transportation terminals and "low energy" and "power assist" doors that are used in less demanding situations. Full documentation of the research activities is available in a two-volume final report. Volume 1 documents the research findings from the state-of-the-art review. Volume 2 provides recommendations with extensive rationale for each.
The objectives of the project were to:
1. Assess the state of the art in human factors research on doors related to accessibility of buildings.
2. Identify the range of automated door products currently available.
3. Investigate constraints related to the use of automated doors.
4. Identify the scope of regulations and standards related to automated doors at the national and state level.
5. Assess international developments in product design and accessibility regulations.
6. Prepare recommendations for revising ADAAG.
3.1 Human Factors of Door Use
The human factors review is organized around a model of the door use process that includes seven subtasks:
1. Perceiving and understanding door operation.
2. Altering gait, adjusting body posture and maneuvering within reach.
3. Reaching and grasping handles, switches or locks.
4. Applying force to overcome resistance of handles, switches or locks.
5. Applying force to overcome resistance of the door, mechanical door closers and pressure differentials.
6. Passing through the doorway, including making adjustments in posture and continuing to apply force.
7. Closing the door and locking it by repeating tasks 1-5 above on the other side of the door.
The abilities of the person, the ambient environment and the social context play a major role in successful completion of the task. They can affect the perception and understanding of door use as well as the level of stress involved in the task. Stress may be related to limitations in ability, difficult environmental conditions such as low levels of illumination or social pressures caused by the presence of others eager to use a door quickly. Automating doors offer an effective way to reduce the stress of door use. Automated doors are specifically designed to reduce congestion and increase access, but they also can be helpful to control access and improve security. Although automated doors can reduce accidents, their mechanical operation, which is outside the control of the individual user, can create potential safety problems. Door use is a critical aspect of safe egress from buildings in emergency situations. Building safety codes and standards reflect this fact through many detailed design criteria. Automated doors must address these emergency concerns if they are part of designated exits.
For people with disabilities, difficulties with door use are more pronounced and often a stressful aspect of everyday experience. Automated doors can make the door use task easier. But, as the analysis and model above make clear, there are many human factors issues that should be addressed in the design of these door systems. Although much research has been completed about door use, there is little research specifically on automated doors. The model of door use can be useful to summarize what is known and to identify the research gaps.
Understanding Door Operation. From existing research, we know that understanding the operation of automated doors can be a problem for people with disabilities and the elderly. We do not know how widespread the problem is since the existing research has observations from only a few subjects. New and innovative products, like automated revolving doors, seem to create the most serious difficulties. Although the existing safety standards require signage on automated doors, we do not know if those provisions are adequate. Furthermore, no research has been done on how information about door operation should be conveyed to visually impaired individuals. Since people who cannot see use their hands and canes to learn about the operation of doors and devices they encounter, attention should also be given to safety for tactile exploration.
Maneuvering. Although research has investigated the need for maneuvering clearances in front of doors, no attention has been given to what clearances might be needed in front of power assisted doors or if the required maneuvering spaces are consistent with automated door safety standards.
Using Door Controls. There has been considerable research on the use of handles, switches and locks by people with disabilities. This research includes specific studies on card slots, push buttons, keys and other devices that are used with locks. Often doors equipped with power operators have high-tech security devices such as card readers. However, there is no research on how to communicate the location and operation of such devices to people with visual impairments.
Force to Operate Controls. Automated doors are activated either through detection systems or manual controls. There is enough research on the use of controls by people with disabilities to make recommendations on size, location and operating forces for these devices.
Force to Open Doors. The accessibility of automated doors is related to the force required to open manual doors. Establishing maximum thresholds for these operating forces essentially determines when automated doors will be required. Much research has been completed on the subject of opening doors against the resistive forces of mechanical closers and air pressure differentials. Although some of the findings are divergent, they can be explained by differences in research methods and sample selection. Given the purpose and intent of an application, it is possible to use the existing data base to make appropriate recommendations for maximum resistance forces (minimum opening forces) at manual doors. Research indicates that the abilities of the more severely disabled population to overcome resistance of door closers are very limited. Closers are not currently designed with a level of efficiency that would allow all doors to close properly if the opening force were set at the limit that people with severe disabilities could manage on an everyday basis. Furthermore, people with severe disabilities have limited use of their hands and arms. Therefore, there is a rationale for requiring automated doors. Only one study has been completed on emergency use of doors (Johnson, 1981). When compared to other studies, the findings indicate that, under emergency conditions, people with disabilities can exert relatively high forces to overcome the resistance of door closers. Thus, there is a rationale for treating doors used only for emergency use or emergency modes of automated doors differently. However, since the documentation of sample selection and findings in that study is not very thorough, it is difficult to evaluate its recommendations.
Passing Through Doorways. Research has demonstrated that passing through doors against the resistance of a closer is quite difficult for many people who use wheelchairs, particularly children. The main problem seems to be that door users have to exert force to keep the door from closing while they are moving through the opening. Safety issues for people who walk or wheel slowly while using automated revolving doors are a special case of this problem. These doors do not really "close", but, the user can be bumped by the leaf behind them. Manufacturers have developed several different approaches to this problem but none has been evaluated in depth.
Closing Doors. Automated doors always close by themselves. Thus, this subtask is not an issue in design.
The samples used in research on automated doors are not fully representative of the disabled community. In particular, very few people with visual impairments have been included. No research has been completed with people who have developmental disabilities or hearing impairments. Controlled laboratory studies have only been conducted with children. In the reports of field research, differences between the people who used the door were not examined in detail. Thus, we have limited information about the variation in abilities among people with different types of disabilities and levels of ability to complete activities of door use.
The research that has been completed addresses many different types of doors and related devices. However, there are some issues that have only been studied with very small samples of individuals. In particular, research using systematic variation of different door features on the same type of door is lacking. The ambient environment has received practically no attention. No research has been completed on the impact of differences in illumination, particularly as it relates to signage. The impact of wind and temperature has not been examined. The impact of crowding is another neglected issue. Finally, little research has been completed on the unique concerns of different building types.
3.2 Products
Our product search identified 86 manufacturers of door-related products -- not all of which necessarily make automated products. Letters of inquiry were sent to all of the manufacturers with a request for information on any automated door products they produce. Of the 86 manufacturers, all but 14 responded. Among these respondents, 18 identified themselves as manufacturers of automated door products and sent catalogs on a total of 121 different product lines including 42 activating devices, 39 operating devices and 40 door systems. All the information obtained was catalogued and described in a database. Each product has been categorized by its manufacturer, type (i.e., activator, operator or door) and defining characteristics. Descriptions of a representative set of product types are included in the report. We interviewed representatives of six companies in depth.
The automated door industry is international in scope. The major manufacturers from the U.S. have interests in Europe. The Japanese market their products in Europe and the U.S. European companies sell their products in the U.S. In some cases, foreign manufacturers produce their products in U.S. plants. One Swedish company makes their entire product in the U.S. except for the circuit boards. We found only one innovative accessibility product made overseas but not in the U.S.
Interviews with six manufacturers suggest that there is not a great deal of technical innovation currently underway. They report that most product development is focused on design improvements such as making the devices smaller and less obtrusive, safer and more secure. In the accessibility area, more attention is being given to wireless remote activation devices. Most companies that produce a large range of different products have specialized devices for accessibility. Other companies specialize only in products to meet accessibility needs. Some products are specifically designed to assist in converting existing manual doors to automated operation.
The manufacturers who produced manual door closers identified a key design problem: code requirements that limit the opening force of exterior doors. The physics behind mechanical door closer design requires the opening force to exceed closing force. As individuals open the door, energy is transferred to the door closer. The closer stores the energy (in a spring or other device) until the door begins to close. The energy stored is then applied to the door through a lever arm. The efficiency of this operation is only about 60%. Thus, an 8.5lbf. (37.8 N) opening force (required by some codes for exterior doors) translates into a 5lbf. (22.2 N) closing force. Manufacturers maintain that such a force is not sufficient to overcome the resistance of door weight, HVAC pressures, wind, gasketing, stiffness of latching mechanisms, installation tolerances, hinge friction, etc. Another problem is that these requirements are difficult to enforce. No one knows what the actual conditions at the site will be. Installation, maintenance and wind behavior, among other considerations, play major roles in determining the actual forces necessary to close the door as well as the force of opening. Such factors cannot be predicted accurately during design and specification of products.
The manufacturers also reported that designers (and possibly code officials) have interpreted the existing codes to mean that low energy and power assisted doors must comply with the manual door opening force requirements so that they can be operated manually when power fails. This results in the development of a product that seems redundant. If a door meets opening force requirements, why would a building owner want to pay extra to automate it? Only facilities that seek ultimate convenience would install such systems.
The market for automated doors is becoming more defined. The type of product that fits best with a particular application is determined by frequency of operation, speed of operation required, new versus existing construction, traffic flow and cost. Products are available to address the full range of applications based on these factors.
Manufacturers believe that automated doors are the preferred means of access for all users, not only people with disabilities. However, when these doors malfunction, users are quick to complain which is not the case with manual doors.
3.3 Regulations and Standards
The two key regulations are the ADA Accessibility Guidelines (ADAAG) and the Uniform Federal Accessibility Standard (UFAS). These are both based in large part on the technical criteria in the national voluntary consensus standard, ANSI A117.1. A new version of the ANSI A117.1 standard has recently been approved (1992). The requirements related to doors in all these documents were identified and compared. With one exception, there are only a few minor differences between them. This is the requirement in ANSI A117.1 (1986) for a maximum opening force at exterior doors of 8.5 lbf. (37.8 N). Neither of the other three documents contains this requirement. Thus, with the advent of ANSI A117.1 (1992) none of the primary sources of accessibility criteria have requirements for opening force.
The states have several different approaches to accessibility regulations. Some states have adopted ANSI A117.1, UFAS, ADAAG or model building code requirements, all of which are relatively similar. Another group of states rely on one or another of these sources as a basis but that have not adopted them by reference. A third, smaller group includes states with their own codes. Severals states have door design criteria that differ from ADAAG. One state, Wisconsin, recommends the use of automated doors where exterior doors have resisting forces greater than 8 lbf. (35.6 N). Two states, Massachusetts and New Hampshire, require automation at exterior doors with resistive forces exceeding 15 lbf. (66.6 N) and interior doors exceeding 8 lbf. (35.6 N). The states of Michigan and Connecticut require that at least one entrance to certain types of buildings have automated doors. The State of Washington recently enacted a change that requires all automated doors to stop and re-open automatically if they encounter a body or object in their path.
Of the 12 Canadian provinces and territories, all either have adopted in total or adapted the Canadian National Building Code. This code requires at least one automated door for certain types of buildings.
The most recent attempt by European countries to develop a consensus standard on accessibility is the European Manual for an Accessible Built Environment sponsored by a committee of the European community. The door design criteria are somewhat different than the U.S. regulations and standards. However, the manual does not require the use of automated doors. The only specific criteria for design of automated doors concerns door speed.
In the review of human factors research (Section 2.1) the findings of research are compared with the requirements of the ADAAG and The European Manual.
3.4 Constraints on the Use of Automated Doors
We completed a survey of nine organizations whose facilities are equipped with automated doors. This survey was completed in two metropolitan area, Buffalo, NY. and Washington, DC. The survey provided insight into issues related to design, installation and use of automated door systems. We also completed a survey of automated door products including interviews with manufacturers.
All the organizations surveyed were generally satisfied with the products that they were using. There were few complaints about the door systems. They are also satisfied with the acquisition costs. Their greatest concern was about reliability and ease of maintenance. Durability, ease of installation, ease of repair, safety and security do not appear to be significant factors in decision making about automated door systems. Energy conservation due to heat loss is a major negative characteristic of automated door systems. However, it is basically a function of high traffic that leads to the need for such systems. Almost all installations surveyed in Buffalo used vestibules to reduce heat loss in winter.
From this survey and our product survey, we identified several constraints related to the use of automated doors: climate, traffic patterns, operating costs, emergency operation and safety.
Climate
Wind loads and pressure differentials, both positive and negative, are design concerns that must be overcome in all exterior door design. Back checking devices stop doors from being damaged when the wind blows against open doors and forces them backwards. Power assist devices can reduce opening force where pressure differentials are required. Energy conservation (heat gain and heat loss) can be addressed through quicker timing of automated doors, or the use of revolving doors, vestibules and air curtains.
High prevailing winds can affect the performance of all types of automated doors. Sliding doors are particularly affected. High winds blowing perpendicular to the doors have a significant impact on the performance of the doors. Yet only a few of the facility managers interviewed reported problems with high winds. This is probably because it is a known factor and is considered in the selection and design of door entries.
The reliability of control mats can be affected by moisture. Snow, ice or heavy rainfall leads to moisture accumulation underneath mats and failure of door controls. Corrosion problems can develop in control mats caused by salt used to melt ice on the sidewalks outside. Cold weather also causes slow operation of exterior pneumatic door operators. In response to such problems, several facility managers in our survey indicated their organizations had abandoned control mats for motion detectors or other sensing systems.
Traffic Patterns
Several organizations reported that pedestrians walking across the front of doors triggered motion detectors. At an airport, even cabs in the loading zone triggered detectors. These problems can be alleviated by adjustments of detection areas. In new construction, the location and orientation of doors can prevent this problem without a reduction in the minimal detection area.
Operating Cost
In general, the cost of operating automated doors themselves is low. The principle operating cost associated with automated doors is due to increased energy costs. Costs increases result not only directly from heat loss or heat gain but also indirectly from additional space, such as vestibules, or equipment, such as air curtains, used to conserve energy at doors with high levels of traffic. However, the energy costs are primarily due to the traffic flow, not the automated doors themselves. Supermarket chains use large vestibules between two sets of doors to help reduce energy costs. One organization said that it had made a decision to utilize only revolving doors in an attempt to cut energy costs. It should be noted that such systems can cost as much as $100,000 each. This commitment indicates the extent of the problem for that particular organization. Low energy door installations do not cause as severe an energy conservation problem because they have less traffic. Although these doors may remain open a long time, several people can pass through before they start closing. This reduces the traffic at adjoining doors.
Although facility managers were satisfied with maintenance costs, the maintenance cost associated with automated door equipment is significant. One organization reported the need to periodically adjust motion detectors. Vandals push them out of alignment. Another facility manager reported that their company (a large supermarket chain) has a full time position devoted to door maintenance. This individual continually visits stores for necessary adjustments and preventative maintenance. Most organizations contacted have a service contract to maintain the doors in working order.
Most organizations contacted are satisfied with the performance of their doors. They are generally considered very reliable. Most report that they have been switching from hydraulic systems to electro-mechanical systems. The later seem to be more reliable and require less maintenance. One organization (a university) reported that they selected sliding doors over swinging doors due to their ease of repair and reliability. For low cost installations, they use swinging doors and maintain them with their own staff. At least six to eight years of trouble-free operation can be expected from new installations. One organization argued that routine maintenance can extend the length of trouble free operation to 20 years, even for pneumatic and hydraulic door systems. They suggested that reports of dissatisfaction are due to a lack of proper maintenance.
In general, the selection of a particular automated door system is based on reliability and durability issues rather then cost concerns. Once the level of reliability and durability is established, cost then becomes an issue for competing systems of the same type.
Emergency Operation
None of the organizations in our facility management survey reported a problem with emergency operation. On the other hand, several seem to have never considered the issue in the past. A few facility managers were prompted by our interview to wonder how their doors would perform under an emergency. They specifically voiced concern about the effect of smoke on motion detectors and infrared detectors. In an emergency when power is curtailed, doors at required exits must operate manually. Hinged, sliding and revolving doors are designed to "break-out" and swing away. However, break-out force must be set high enough to prevent this from happening during everyday operation.
Safety
In general the organizations we surveyed were satisfied with the physical safety and security of automated door products for public use under general operating conditions. Only a few minor security problems were mentioned related to vandalism. There were some serious safety problems reported, however. It was noted that children can slip under the guardrails and step on control mats causing the door to open. Two supermarket chains with large numbers of doors both reported incidents. One had a number of incidents in which doors closed on the fingers of small children who inserted them along the hinged edge of the door. They now have installed guards along the hinge to prevent the doors from closing when an object is detected along that edge.
3.5 Trends in Product Development
As described above in 3.1, the key problem in use of mechanical closers is providing enough force to overcome resistance to door closing. Since the design margins are small to begin with, there is little tolerance to work with. Compliance with an 8.5 lbf. (37.8 N.) maximum opening for opening doors thus cannot be assured until after construction or installation. The closing force required to ensure proper latching may require greater opening forces than this maximum limit.
There is no new technology being developed related to mechanical door closers that would alleviate the need for automated doors as an accessibility feature. The primary design problem for mechanical closers is that a reduced opening force means a reduced closing force. Since it always takes more force to open the door than close it, the lower the allowable limit on the opening force, the more difficult it is to get the door to close (i.e., higher efficiency). Lower closing force is a serious product liability issue for the closer manufacturer, especially as it relates to security.
In general the most common types of automated doors installed are swinging doors with electro-mechanical, pneumatic or hydraulic operators. Doors that are installed primarily to facilitate accessibility for people with disabilities are usually activated by a touch/pressure switch although some products have a feature that activates power upon pushing the door. One organization surveyed had discovered, through statistical analysis, that their hydraulic and pneumatic doors actually had a longer trouble-free life expectancy, but a rigorous routine maintenance schedule had to be followed. Thus, the trend toward electro-mechanical systems may be due to a desire to reduce scheduled and preventive maintenance rather than improving reliability.
There is little that any manufacturer is willing to say about the development of new technology in automated doors. The focus of research and development seems to be in the areas of reduced sizes of components, aesthetics, activation devices, improved safety systems, and security improvements. The manufacturers as a group are most concerned about safety first.
There is a growing interest in the use of wireless remote units for installations. One idea would have the Federal Communications Commission provide reserved radio frequencies for all manufacturers of automatic doors and door openers. People with disabilities could receive a single small transmitter/receiver that they could then use through this universal protocol. This would enable the user to receive audio messages upon approach to a facility with an automatic or power assisted entry door, instructions as to its location, warning upon approach, and the means to activate the door. Such a system would be most beneficial to persons with visual impairments.
Three other innovative technologies identified include sensors that prevent revolving door wings from hitting a user, an add-on power wheel that rolls a door shut and smart technology that can recognize the difference between the wind and solid obstacles in the door path. The automated revolving door is a new technology that addresses a major short-coming of other automated door systems, poor performance with respect to energy conservation.
Although installation of automated doors can improve accessibility significantly to the point that all building users notice, existing buildings do present some difficulties in installation. Products that do not attach permanently to the door (i.e. "retrofit" applications) have some benefit related to historic property concerns. Apparently this is more acceptable than permanent alterations, even though the door opener is still a visual intrusion.
4.1 Performance Criteria
The findings of our research were used to develop the following performance criteria for accessible automated doors.
4.1.1. Required Automated Doors: All new buildings used by the public should have at least one automated door at an accessible entrance. There should be an exception for small buildings where adding such a door may be a financial hardship for building owners.
4.1.2. Automated Revolving Doors and Turnstiles: Where automated revolving doors are used at an accessible entrance, an alternative accessible means of access should also be available at the entrance.
4.1.3. Security Controls: Keyed switches, card readers or combination switches should be allowed as security devices at required automated doors. They should be used only at entrances that are kept locked and where access is restricted.
4.1.4. Door Width: Automated doors must be wide enough for use with wheeled mobility devices, walking aids or by very large individuals. Full powered or low energy bi-parting and telescoping doors should be allowed to meet this requirement based on the width of the entire opening rather than one panel, as with manual doors.
4.1.5 Door Width -- Emergency Use: Automated doors should be wide enough for emergency egress and there should be an alternative accessible means of egress.
4.1.6. Thresholds and Edges: Thresholds and control mats at automated doors should not have abrupt edges that would pose a barrier or safety hazard to those who use wheeled mobility devices or walking aids or those who are visually impaired.
4.1.7. Compartments in Revolving Doors: Automated revolving doors should have enough space within them for use of wheeled mobility devices and walking aids.
4.1.8. Door Timing: Automated doors should remain open long enough to allow people with disabilities to enter the opening and pass through the door.
4.1.9. Opening Force: The force required to open a manual door or an automated door when the building power is off should be limited to ensure that people with disabilities can escape a building during an emergency. An alternative is a back-up power supply to keep the door operating.
4.1.10. Bump Force: The force produced by a low energy door as it closes or opens should be limited to avoid knocking an individual off balance. Two options should be allowed, a maximum force threshold and sensor controlled variable forces.
4.1.11. Door Swing: Full powered doors that swing against the direction of travel should have protective features to ensure that they will not hit someone approaching the door.
4.1.12 Activating Controls: Controls switches should be easy to use by people who have difficulty forming a grip.
4.1.13. Activation Forces: Forces necessary to operate door controls should be within the capabilities of severely disabled people.
4.1.14 Control Location: Switches for operating low energy automated doors should be located as follows:
a. Within the reach range of people with severe disabilities who use wheeled mobility devices.
b. In a location that allows a direct approach to the door.
c. In close proximity to the door.
d. In standard locations.
f. Not on the door itself.
4.1.15. Detection Zone: Sensors and control mats at the pull side of hinged doors should detect people approaching doors early enough to ensure that the door will open before the user reaches the sweep area.
4.1.16. Visual Instructions and Warnings: Warning signs should be in highly visible locations, have standardized symbols and be large enough to be read by people with visual impairments. Switches for low energy automated doors should be identified with the International Symbol of Accessibility.
4.1.17 Tactile Information: Tactile information should be provided to help visually impaired people become aware of low powered automated doors and help them find the location of switches.
4.1.18. Maneuvering Space: There should be enough maneuvering space in front of automated doors and controls to accommodate use of wheeled mobility devices.
4.1.19. Ground and Floor Surfaces: Floor and ground surfaces in the maneuvering clearances and at the control location should not have a slope exceeding the minimum required for drainage.
4.1.20. Background Noise: If audible warnings or instruction messages are provided; they should be distinguishable against the ambient background noise and be accompanied by visual warnings or instruction labels.
4.2 Technical Criteria
Specific recommendations for revising the ADAAG technical criteria based on the performance criteria above were then developed. The recommendations are as follows (ADAAG Paragraph numbers are in parentheses):
a. Change the number of this paragraph to 4.13.3, as described above. Update references:
Full powered automated doors: comply with ANSI/BHMA A156.10 -1991
Low energy and power assist doors: comply with ANSI/BHMA A15619 - 1990
b. Delete the existing ADAAG requirements for door timing and "bump force."
c. Add a sentence as follows:
Where requirements of ADAAG differ from those of ANSI/BHMA A156.10 - 1991 and ANSI/BHMA A156.19 - 1990, then those of ADAAG shall be followed.
4.2.2 Revolving Doors (new section)
a. An automated revolving door may be used at an accessible doorway. If such a door is used, an accessible swinging or sliding door shall be located to serve the same space and facilitate the same pattern of use. This alternative accessible door may be a manual door.
b. Accessible automated revolving doors shall comply with all the applicable requirements of Section 4.13.3, Automated Doors.
c. Accessible automated revolving doors shall be identified by the International Symbol of Accessibility, as in 4.30.7.
d. Automated turnstiles shall not be allowed as part of an accessible path of travel. Where an automated turnstile is provided, an accessible door or gate shall be located to serve the same space and facilitate the same pattern of use. This alternative accessible door or gate may be manually operated.
4.2.3 Gates (new section)
Automated gates shall meet all applicable specifications of 4.13.3 when they are used at accessible entrances.
4.2.4 Doorways with Multiple Panels (new section)
The minimum clear width for automated door systems with multiple door panels that open simultaneously shall be based on the clear opening provided by all panels in the open position.
4.2.5 Clear Opening Width (new section)
a. The minimum clear width of door openings shall be 32 in. (815 mm.) in power-on mode.
b. The minimum clear width of door openings shall be 32 in. (815 mm) in power-off mode, or, an accessible door that also meets the requirements for a means of egress shall be provided immediately adjacent to the automated door.
c. If an automated revolving door is used in an accessible path of travel each compartment shall be large enough to accommodate a person in a wheelchair.
4.2.6 Maneuvering Clearances (new section)
a. Minimum maneuvering clearances at automated revolving doors shall be as follows:
1) Approach clearance: a clear floor space 48 in. (1220 mm.) deep and as wide as the door opening.
2) Control clearance: a clear floor space 48 in. (1220 mm.) deep extending at least 24 in. (610 mm.) beyond the edge of the doorway.
b. Minimum maneuvering clearances at sliding and swinging doors shall be as follows:
1) Front approach:
Push-side approach: a clear floor space as wide as the doorway and 48 in. (1220 mm.) deep measured from the face of the door in a closed position. This area may overlap a control mat or detection area.
Pull-side approach: a clear floor space as wide as the doorway and 48 in. (1220 mm.) deep starting beyond the swing of the door. This area may overlap a control mat or detection area.
2) Side approaches :
Length: a clear floor space as wide as the doorway plus 48 in. (1220 mm.) measured from the near edge of the door frame. This area may overlap a control mat or detection area.
Width: a clear floor space at least 60 in. (1525 mm.) for latch side approaches, 72 in. (2100 mm.) for hinge side approaches. This area may overlap a control mat or detection area.
Door swing clearance for hinge side approaches: at least 36 in. (915 mm.). This area may overlap a control mat or detection area.
Exception:
Clearances at power assisted doors do not have to comply with these requirements; however, they shall comply with the requirements for clearances at manual doors as in 4.13.2. 6.
Note:
Minimum maneuvering clearances for required automated doors are not intended for design of workplace and home modifications. In these situations, clearances should be based on the abilities of individual workers and residents and/or the performance of their wheelchairs.
4.2.7 Two Doors in Series (new section)
a. The minimum space between two doors in series: 48 in. (1220 mm) plus the width of any door swinging into the space.
b. Doors in series shall swing either in the same direction or away from the space in between.
c. If doors are not operated by automatic detection devices, controls for at least one door shall be located in the vestibule.
d. A single automated detection device or switch is allowed to control both doors either from outside a vestibule or from within, as long as controls are available as in paragraph c. Where such control is provided, the opening cycle of the second door shall have a time delay equal to one second for each foot (30 mm) of space between the two doors.
4.2.8 Thresholds and Floor Mats
a. Maximum threshold height at automated doors: 1/2 in. (13 mm.).
b. Edges of floor mats used in activation and safety areas of automatic doors: 1/4 in. (6 mm.) maximum or beveled with slope no greater than 1:4
4.2.9 Automatic Detection and Safety Systems (new section)
a. Automatic detection and safety systems shall be provided for all full powered swinging and sliding automated doors as required by ANSI/BHMA A156.10.
b. Automated revolving doors shall have an automatic detection and safety system that stops the door without contact upon detection of a stationary object or person within the compartments of the door or within the door opening.
4.2.10 Door Timing (new section)
a. Low energy doors shall have a hold open period and closing period as required by ANSI/BHMA A156.19 or be equipped with an automated detection and safety system. Such a system shall sense the approach of an individual, and open the door; if the door encounters a person or object in its path while opening or closing, it must stop and reverse direction without contact.
b. Automated revolving doors shall have a slow mode that reduces the speed of the door to two revolutions per minute, maximum. A sensing system that adjusts door speed automatically to the speed of door users is an acceptable method to comply with this requirement.
4.2.11 Door Operating Forces (new section)
a. Low energy doors
ALT. 1: The force a closing door exerts on a person or object within its path shall be 15 lbf. (67 N.) maximum.
ALT. 2: The force a closing door exerts on a person or object within its path shall be 15 lbf. (67 N.) maximum. A sensing system shall also be provided that immediately stops the door upon contact and automatically reverses its direction.
ALT. 3: A sensing system shall be provided that detects an object in the path of the door, stops and reverses its direction without contact.
Note: The closing force shall be measured while the door is stopped to avoid momentary and inconsistent readings. This is the same approach used in the ANSI/BHMA A156 Standards. The opening force requirements for manual doors do not apply to low energy doors used in power-off mode.
b. Power assist doors shall have opening forces no greater than 5 lbf. (22.2 N.). Forces shall be measured at the handle or 30 in. (760 mm.) from the hinge if the door has a panic bar or has no operating hardware.
Note: Opening forces shall be measured with the power assist on. The opening force requirements for manual doors do not apply to power assist doors in manual mode.
c. Breakout forces for all automated doors under emergency conditions shall comply with one of the following options:
1) 25 lbf. (111.2 N.) maximum
2) 50 lbf. (222.4 N.) max. and emergency backup power supplied by a battery or an emergency power system that will operate the door for at least one hour in the event of power failure.
4.2.12 Control Switches (new section)
a. If an automated door is not activated by an automatic sensing device or control mat, switches shall be located at each approach to the door. Switches shall be either push buttons, push plates or a detector that can be activated by a hand movement.
b. Push buttons and plates shall have a minimum width of .75 in. (20 mm.). The maximum force to activate a push button or plate shall be 3lbf. (13.3 N). If a push plate is 3 in. (75 mm.) wide or larger, the maximum force of activation may be increased to 5 lbf. (22.2 N.). In facilities for children, the maximum force of activation shall be 2 lbf. (8.9 N.) for all buttons and plates, regardless of size.
c. The usable surface of switches shall not be recessed below the surface in which they are installed.
d. Security devices such as keyed locks, keypads, card readers and swipes are allowed to limit access to door controls at entrances where doors are kept locked. Slots shall be at least 1/8 in. (3 mm.) wide or have guides wider than 1/8 in. (3 mm.) apart directing the card into the slot. Grip clearances around the card reader shall be at least 6 in.(150 mm.).
e. Self activation through movement of the door is allowed only on one way doors that are pushed to open and have no latch.
f. Control switches shall be located as follows:
Mounted no higher than 36 in. (915 mm.) on center, measured from the floor.
With a wheelchair maneuvering clearance.
Outside the arc of the door swing.
No more than 72 in. (1830 mm.) from the door it operates.
For side approaches on the pull side, positioned opposite the door.
4.2.13 Signage (new section)
a. Labels and warning for all automated doors shall be provided as required in ANSI/BHMA A156.10.
b. Instructional signage shall comply with 4.30.2 and 4.30.5.
c. Controls for automated doors shall be identified by the International Symbol of Accessibility (4.30.7). The symbol shall be located on the switch or immediately adjacent to it.
Wherever audible warning sounds and messages are provided, comparable signs or illuminated warning signals shall also be provided.
Where two adjacent doors are each designated for one way traffic, the stream of traffic in the direction of the door shall be on the right.
Detailed rationales for each recommendation were included, including cost implications. Several Figures were developed to illustrate the requirements.
Automated door products are a significant aid to accessibility. Many groups of people with disabilities can benefit significantly from the installation of these products at building entrances. Not only do they assist people with the most severe disabilities, those who use electric wheelchairs and have limited use of their arms and hands, but, they also provide assistance to the frail elderly with limited strength to open a manual door with a mechanical closer. In fact, automated doors are clearly a great benefit to the general population as well. They are a good example of universal design.
Our review of products and survey of existing applications demonstrated that there are many reliable products on the market. Technology is improving steadily at both the high end and low end. Although automated doors require more maintenance and service than manual doors, this is partly due to the fact that when they do not operate properly, it is more noticeable and has a greater impact on the building user than a malfunctioning manual door does. Full powered automatic doors are available in a wide range of options to suit the needs of all applications where there is a high volume of pedestrian traffic. Low energy automated doors provide accessibility at a relatively low cost. Low energy "retrofit" controls can be added to existing manual doors with ease.
The human factors research on door use by people with disabilities clearly demonstrates a need for requiring automated doors as part of accessibility standards and regulations. Some state codes and the Canadian National Building Code already require automated doors for some types of buildings. However, a human factors analysis also demonstrates that there are many design issues related to the accessibility and safety of automated door systems. Automated door control systems, in particular, must be designed with the capabilities and anthropometrics of the users in mind. Although automated doors are similar in many respects to manual doors, in others they are quite different, not only in how they operate, but also in who uses them. Some existing accessibility criteria for manual doors, for example, maneuvering clearances, have to be re-considered for automated doors. Although the existing consensus standards for automated doors address many safety concerns, they do not take into account some important accessibility issues and also do not cover some types of doors, such as automated revolving doors. Therefore, it is necessary to develop accessibility standards to ensure that these products will be usable for people with disabilities.
Through an analysis of the door use task, this research identified accessibility requirements and specific recommendations for improving the coverage of automated doors in accessibility standards and regulations. The recommendations are prepared in a way that makes them easy to add to the ADAAG document. Each one is accompanied by a detailed rationale with cost implications for building design and operation. The research documented in this report will assist those who write standards and codes in improving the coverage that those documents give to automated door products.
Although there was much information available for making informed recommendations, there are some gaps in the knowledge base about automated doors and doors in general. The areas where additional research would be useful include:
1. Door maneuvering clearances for scooters.
2. Door opening force for manual doors.
3. Use of automated revolving doors and turnstiles.
4. Use of dual panel swinging doors.
5. The impact of "bump" force on door use.
6. Use of remote controls.
7. Identification of controls for people with visual impairments.
8. Usability of ANSI/BHMA A156 signage requirements.
Preface
This is the first volume of the final report for the Automated Doors Project. This volume documents a state of
the art review on the subject.
Many dimensions are provided in this report. To aid readers from all countries, we have provided both English and Metric units. Our approach to metric conversion can be termed eclectic. ADAAG, UFAS and the ANSI A117.1 (1986) Standard use a round-off approach to metric conversion. Each metric equivalent is rounded off to the nearest 5 mm since exact conversion to the millimeter is not meaningful in most construction dimensions. The BHMA/ANSI Standards for automated doors, however, use a different approach. Research results from countries outside the U.S. are reported in metric dimensions. Moreover, American research is usually not reported in metric. Thus, in reporting research results we used exact conversions to avoid the appearance of mistakes, but in reporting on standards, we used the conversions as found in those referenced. Although this might cause some confusion, it is less than what would be caused if we used one method consistently since many readers would wonder why our conversions are different then those in familiar sources.
Please note that the ADAAG omits the units from dimensions on illustrations. The English units in inches are above the dimension line and the metric units in millimeters are below. Illustrations in this report taken from ADAAG follow this convention.
The research and publication of this report were made possible through a research contract from the U.S. Department of Education, Contract No. QA 92005001. The contents of this report do not necessarily reflect the view of the Department or the U.S. Access Board.
1. 0 Introduction
This report documents a literature review, product review, facility management survey and code review of automated doors. Automated doors include full powered automated doors such as those that are commonly used at supermarkets and transportation terminals and "low energy" and "power assist" doors that are used in less demanding situations. The focus of this review is accessibility for people with disabilities.
The objectives of the report are to:
Four research tasks were completed. They included:
1.3.1 Human Factors of Door Use
Based on the work of other researchers, a model of the door use process was developed that includes seven subtasks:
The abilities of the person, the ambient environment and the social context play a major role in successful completion of the task and can affect the perception and understanding of door use as well as the level of stress involved in the task. Stress may be related to limitations in ability and difficult environmental conditions, such as low levels of illumination or social pressures. Automating doors reduces the stress of door use. Automated doors are specifically designed to reduce congestion and increase access, but they also can be helpful to control access and improve security. Although automated doors can reduce accidents, their mechanical operation, which is outside the control of the individual user, also can create potential safety problems. Door use is also a critical aspect of safe egress from buildings in emergency situations. Building safety codes and standards reflect this fact through many detailed design criteria. Automated doors must address these emergency concerns if they are part of designated exits.
For people with disabilities, difficulties with door use are more pronounced and often a stressful aspect of everyday experience. Automated doors can provide many benefits to people with disabilities by making the door use tasks easier. But, as the analysis and model above make clear, many human factors issues should be addressed in the design of these door systems. Although much research has been completed about door use, there is little research specifically on automated doors. The model of door use can be useful to summarize what we do know and identify the research gaps.
From existing research, we know that understanding the operation of automated doors can be a problem for people with disabilities and the elderly. We do not know how widespread the problem is since the existing research contains observations from only a few subjects. New and innovative products, such as automated revolving doors, seem to create the most serious difficulties. Although the existing safety standards require signage on automated doors, we do not know if those provisions are adequate. Furthermore, no research has been done on how information about door operation should be conveyed to visually impaired individuals. Since people who cannot see use their hands and canes to learn about the operation of doors and devices they encounter, attention should also be given to safety for tactile exploration.
Considerable research has been completed on the use of handles, switches and locks by people with disabilities. This research includes specific studies on card slots, push buttons, keys and other devices that are used with locks. Often doors equipped with power operators have high-tech security devices such as card readers, but no research has been completed on how to communicate the location and operation of such devices to people with visual impairments.
Although considerable research has investigated the need for maneuvering clearances in front of doors, no attention has been given to what clearances are needed in front of power assisted doors or if the required maneuvering spaces are consistent with automated door safety standards.
From the perspective of accessibility, the specification of automated doors is tied to the force required to open doors. If a door handle, lock or door presents too great a resistive force for people with disabilities to overcome, then an automated system can be specified to substitute. Much research has been completed on the subject of opening doors against the resistive forces of mechanical closers and air pressure differentials. Although some of the findings are divergent, they can be explained by differences in research methods and sample selection. Given the purpose and intent of an application, it should be possible to use the existing data base to make appropriate recommendations for maximum resistance forces (minimum opening forces) of manual doors.
Research indicates that the abilities of the more severely disabled population to resist forces of doors closers are very limited. Closers are not currently designed with a level of efficiency that allows all doors to close properly if the opening force is set at the limit that this group of people could manage on an everyday basis. Furthermore, many people with severe disabilities have limited use of their hands and arms. Therefore, there is a rationale for requiring automated doors. Only one study has been completed on emergency use of doors (Johnson, 1981). When compared to other studies, the findings indicate that, under emergency conditions, people with disabilities can exert relatively high forces to overcome the resistance of door closers. Thus, there is a rationale for treating doors used only for emergency use or emergency modes of automated doors differently. However, since the documentation of sample selection and findings in that study is not very thorough, it is difficult to evaluate its recommendations.
Research has demonstrated that passing through doors against the resistance of a closer is quite difficult for many people who use wheelchairs, particularly children. The main problem seems to be that door users have to exert force to keep the door from closing while they are moving through the opening. Safety issues for people who walk or wheel slowly while using automated revolving doors are a special case of this problem. These doors do not really "close", but, the user can be bumped by the leaf behind them. Manufacturers have developed several different approaches to this problem but none has been evaluated in depth.
The samples used in research on automated doors are not fully representative of the disabled community. In particular, very few people with visual impairments were included. No research has been completed with people with developmental disabilities or hearing impairments. Controlled laboratory studies have only been conducted with children. In the reports of field research, differences between the people who used the door were not examined in detail. Thus, we have limited information about the variation in abilities among people with different types of disabilities and their levels of ability to complete activities of daily living tasks.
The research that has been completed addresses many different types of doors and related devices. However, some issues have only been studied with very small samples of individuals. In particular, systematic variation of different door features on the same type of door is lacking. The ambient environment has received practically no attention. No research has been completed on the impact of differences in illumination, particularly as it relates to signage. The impact of wind and temperature has not been examined. The impact of crowding is another neglected issue. Finally, little research has been completed on the unique concerns of different building types.
1.3.2 Products
Our product search identified 86 manufacturers of door-related products -- not all of which necessarily make automated products. Letters of inquiry were sent to all of the manufacturers with a request for information on their automated door products, if they produce them. Of the 86 manufacturers, all but 14 responded. Among these respondents, 18 identified themselves as manufacturers of automated door products and sent catalogs on a total of 121 different product lines including 42 activating devices, 39 operating devices and 40 door systems. All the information obtained was catalogued and described in a database. Each product has been categorized by its manufacturer, type (i.e., activator, operator or door) and defining characteristics. Descriptions of a representative set of product types are included in the report. We interviewed representatives of six companies in depth.
The automated door industry is international in scope. The major manufacturers from the U.S. have interests in Europe. The Japanese market their products in Europe and the U.S. European companies sell their products in the U.S. In some cases, foreign manufacturers produce their products in U.S. plants. One Swedish company makes its entire product in the U.S. except for the circuit boards. We found only one innovative accessibility product made overseas but not in the U.S., a low cost, battery operated device for residential use.
Interviews with six manufacturers suggest that there is not a great deal of technical innovation currently underway. The manufacturers report that most product development is focused on design improvements like making the devices smaller and less obtrusive, safer and more secure. In the accessibility area, more attention is being given to wireless remote activation devices. Most companies that produce a large range of different products have specialized devices for accessibility. Other companies specialize only in products to meet accessibility needs. Some products are specifically designed to assist in converting existing manual doors to automated operation.
The manufacturers of manual door closers identified a key design problem -- meeting the requirements of accessibility codes that have limitations on the opening force of exterior doors. The physics behind mechanical door closer design requires the opening force to exceed closing force. As individuals open the door, they transfer energy to the door closer. The closer stores the energy (in a spring or other device) until the door begins to close. The energy stored is then applied to the door through a lever arm. The efficiency of this operation is only about 60 percent. Thus, an 8.5 lbf. (37.8 N.) opening force (required by some codes for exterior doors) translates into a 5 lbf. (22.2 N.) closing force. Manufacturers maintain that such a force is not sufficient to overcome the resistance of door weight, HVAC pressures, wind, gasketing, stiffness of latching mechanisms, installation tolerances, hinge friction, etc. Another problem is that these requirements are difficult to enforce. No one knows the actual site conditions. Installation, maintenance and wind behavior, among other considerations, play major roles in determining the actual forces necessary to close the door as well as the force of opening. Such factors cannot be predicted accurately during design and specification of products.
The manufacturers also reported that designers (and possibly code officials) have interpreted the existing codes to mean that low energy and power assisted doors must comply with the manual door opening force requirements when they are operated manually such as when power fails. This results in the development of a product that seems redundant. If a door meets opening force requirements for manual doors, why would a building owner want to pay extra to automate it? Only facilities that seek ultimate convenience would install such systems.
The market for automated doors is becoming more defined. The type of product that fits best with a particular application is determined by frequency of operation, speed of operation required, new versus existing construction, traffic flow and cost. Products are available that address the full range of applications based on these factors.
Manufacturers believe that automated doors are the preferred means of access for all users, not only people with disabilities. Automated door malfunctioning, however, is more critical than the malfunctioning of manual doors.
1.3.3 Regulations and Standards
ADA Accessibility Guidelines (ADAAG) and the Uniform Federal Accessibility Standard (UFAS) are the two key regulations that govern accessibility to automated doors. These regulations are based in large part on the technical criteria in the national voluntary consensus standard, ANSI A117.1. A new version of the ANSI A117.1 standard has recently been approved (1992). The requirements related to doors in all these documents were identified and compared; with one exception, only a few minor differences exist between them. The one difference is the requirement in ANSI A117.1 (1986) for a maximum opening force at exterior doors of 8.5 lbf. (37.8 N.). Neither of the other three documents contains this requirement. Thus, with the advent of ANSI A117.1 (1992), none of the primary sources of accessibility criteria have requirements for opening force.
Our survey of state codes demonstrated great diversity in how they adopt accessibility regulations. Some states adopt ANSI A117.1, UFAS, ADAAG or model building code requirements, all of which are relatively similar. The codes of another group of states rely on one or another of these sources as a basis but don't adopt them by reference. A third, smaller group have their own codes. Several states have door design criteria that differ from ADAAG's. One state, Wisconsin, recommends the use of automated doors with exterior doors with resisting forces greater than 8 lbf. (35.6 N.). Two states, Massachusetts and New Hampshire, require automation at exterior doors with resistive forces exceeding 15 bf. (66.6 N.) and interior doors exceeding 8 lbf. (35.6 N.). The states of Michigan and Connecticut require that at least one entrance to certain types of buildings have automated doors. Washington State recently enacted a change that requires all automated doors to stop and re-open automatically if they encounter a body or object in their path.
Of the 12 Canadian provinces and territories, all either adopt in total or adapt the Canadian National Building Code that code requires at least one automated door for certain types of buildings.
The most recent attempt by European countries to develop a consensus standard on accessibility is The European Manual for an Accessible Built Environment sponsored by a committee of the European Community. The door design criteria are somewhat different than the U.S. regulations and standards. However, the manual has no criteria to mandate automated doors. The only specific criteria for design of automated doors concerns the speed of the door.
In the review of human factors research (Section 2.1) the findings of research are compared with the requirements of the ADAAG and The European Manual.
We completed a survey of nine organizations whose facilities are equipped with automated doors. This survey was completed in two metropolitan areas -- Buffalo, NY and Washington, DC. The survey provided insight into issues related to design, installation and use of automated door systems. We also completed a survey of automated door products, including interviews with manufacturers.
All the organizations surveyed were generally satisfied with the products they were using. We noted few complaints about the door systems. The organizations were also satisfied with the acquisition costs. Their greatest concern was about reliability and ease of maintenance. Ease of installation, ease of repair, safety and security do not appear to be significant factors in decision making about automated door systems. Energy conservation due to heat loss is a major negative characteristic of automated door systems. However, it is basically a function of high traffic that leads to the need for such systems. Almost all installations surveyed in Buffalo used vestibules to reduce heat loss in winter.
Wind loads and pressure differentials, both positive and negative, are design concerns that must be overcome in all exterior door design. Back checking devices stop doors from being damaged when the wind blows against open doors and forces them backwards. Power assist devices can reduce opening force where pressure differentials are required. Energy conservation (heat/cool gain and heat/cool loss) can be addressed through quicker timing on automated doors, revolving doors, vestibules and air curtains.
High prevailing winds can affect the performance of all types of automated doors. Sliding doors are particularly affected. High winds blowing perpendicular to the doors have a significant impact on the doors' performance. Yet only a few of the facility managers interviewed reported problems with high winds. This is probably because it is a known factor and is considered in the selection and design of door entries.
The reliability of control mats can be affected by moisture. Snow, ice or heavy rainfall lead to moisture accumulation underneath mats the subsequent failure of door controls. Corrosion problems can develop in control mats caused by salt used to melt ice on the sidewalks outside. Cold weather also causes slow operation of exterior pneumatic door operators. In response to such problems, several facility managers in our survey indicated their organizations had abandoned control mats for motion detectors or other sensing systems.
Motion detectors created unanticipated problems for several organizations; pedestrians walking across the front of the opening can trigger the detector. At an airport, cabs in the loading zone triggered the detector. These problems can be alleviated by adjustments of detection areas. In new buildings, location and orientation of doors can solve the problem without resorting to a minimal detection area.
The facility managers surveyed were satisfied with the purchase cost of automated doors and generally with the cost of maintenance. Swinging doors are less costly than sliding doors. In general, the cost of operating the doors themselves is low. The principle operating cost associated with automated door is due to energy loss. Energy costs increase due to heat escaping out the doors, but also with the need for additional space, such as vestibules or equipment such as air curtains that are often needed to alleviate such problems in installations with high traffic. Energy losses are primarily due to the traffic flow, not automated doors themselves. Supermarket chains use large vestibules between two sets of doors to help reduce energy costs. One organization said that it had made a decision to utilize only revolving doors in an attempt to cut energy costs. It should be noted that such systems can cost as much as $100,000 each. This commitment then indicates the extent of the problem for that particular organization. Low energy door installations do not cause as severe an energy problem because they have less traffic. Although these doors may remain open a long time, several people can pass through before they start closing, reducing the traffic at other doors.
Significant maintenance costs are associated with automated door equipment. One organization reported the need to periodically adjust motion detectors. Vandals push them out of alignment. One facility manager reported that his company (a large supermarket chain) has a full-time employee devoted to door maintenance who continually visits stores and makes necessary adjustments and perform preventative maintenance tasks. Most organizations contacted have a service contract to maintain the doors in working order.
Most organizations are satisfied with the performance of their doors, which are generally considered very reliable. Most report that they have been switching from hydraulic systems to electro-mechanical systems. The later seem to be more reliable and require less maintenance. One organization (a university) reported that it selected sliding doors over swinging doors because of their ease of repair and reliability. For low cost installations, they use swinging doors and maintain them with their own staff.
At least six to eight years of trouble-free operation can be expected from new installations. One organization argued that routine maintenance can extend the length of trouble free operation to 20 years, even for pneumatic and hydraulic door systems. The organizatoin suggested that reports of dissatisfaction are due to absence or inattention to scheduled maintenance programs.
In general, the selection of a particular automated door system is based on reliability and durability issues, rather then cost concerns. Once reliability and durability levels are established, cost becomes an issue for competing systems of the same type.
None of the organizations in our facility management survey reported a problem with emergency operation. In fact, some seem to have never considered the issue. A few facility managers were prompted by our interview to wonder how their doors would perform in an emergency. They specifically voiced concern about the effect of smoke on motion detectors and infrared detectors. In an emergency, when power is curtailed, doors at required exists must operate manually. Hinged, sliding and revolving doors are designed to "break-out" and swing away. However, break-out force must be set high enough to prevent this from happening during everyday operation.
As described above, the key problem with using closers is providing enough force to overcome resistance to door closing. Since so many factors relate to field conditions, one often does not know what the actual resistance forces will be until the door is installed. Since the design margins are small to begin with, there is little tolerance to work with. Compliance with a criterion cannot be assured until after the fact for exterior doors. The closing force required to ensure proper latching may require greater opening forces.
No new technology related to mechanical door closers is being developed that would alleviate the need for automated doors as an accessibility feature. The primary design problem for mechanical closers is that a reduced opening force means a reduced closing force. Since it always takes more force to open the door than close it, the lower the allowable limit on the opening force, the more difficult it is to get the door to close (i.e. requires higher efficiency if possible). This is a serious product liability issue for the closer manufacturer, especially as it relates to security.
In general the most common types of automated doors installed are swinging doors with electro-mechanical, pneumatic or hydraulic operators. Doors that are installed primarily to facilitate accessibility for people with disabilities are usually activated by a touch/pressure switch, although some products have a feature that activates power upon pushing the door.
There is little that any manufacturer is willing to say about the development of new technology in automated doors. The focus of research and development seems to be in the areas of reduced sizes of components, aesthetics, activation devices, improved safety systems, and security improvements. The manufacturers as a group are most concerned about safety first.
There is a growing interest in the use of wireless remote units for installations. One idea would have the FCC provide a reserved radio frequencies for all manufacturers of automatic doors and door openers. People with disabilities could receive a single small transmitter/receiver that they could then use through this universal protocol. This would enable the user to receive audio messages upon approach to a facility with an automatic or power assisted entry door, information about its location, warning upon approach and how to activate the door. Such a system would be most beneficial to persons with visual impairments.
Three other innovative technologies identified include sensors that prevent revolving door wings from hitting a user, an add-on power wheel that rolls a door shut and smart technology that can recognize the difference between the wind and solid obstacles in the door path.
The automated revolving door was mentioned as a new technology that addresses a major short-coming of other automated door systems, namely poor performance with respect to energy conservation counter to the trend.
One organization surveyed had discovered through statistical analysis, that their hydraulic and pneumatic doors actually had a longer trouble-free life expectancy, but a rigorous routine maintenance schedule had to be followed. Thus, the trend toward electro-mechanical systems may be due to a desire to reduce scheduled and preventive maintenance rather than improve reliability.
Although installation of automated doors can improve accessibility significantly to the point that all building users notice, existing buildings present some difficulties in installation. Products that do not attach permanently to the door have some benefit related to historic property concerns. Apparently this is more acceptable than permanent alterations, even though the door opener is still a visual intrusion.
Once installed, automatic and power assisted doors tend to become the preferred means of access for all users, not only people with disabilities. Also, when these doors do not function up to 100 percent capability, the users are quick to complain. This is in contrast to generally acceptable problems that might be experienced with manual doors. One organization (a convention center) reported that automatic doors installed for accessibility purposes were the most popular for all building users.
There are some minor problems fitting automated door systems into existing buildings. A major federal performing arts facility in Washington, DC, had a custom-designed high quality finish with a bronze and white color motif. New automated door systems installed in the building were anodized aluminum and clashed with the existing aesthetics. A school district reported that an unanticipated cost for their installation was due to the difficult wall patching and electrical costs related to the type of construction in their building (masonry).
In general the organizations surveyed were satisfied with the physical safety and security of automated door products for public use under general operating conditions. Only a few minor security problems were mentioned related to vandalism. There were some serious safety problems reported, however. It was noted that children can slip under the guardrails and step on control mats causing the door to open. Two supermarket chains with large numbers of doors both reported incidents. One had a number of incidents in which doors closed on the fingers of small children who inserted them along the hinged edge of the door. They now have installed guards along the hinge to prevent the doors from closing when an object is detected along that edge.
2.0 Review Section
2.1.1 Search Process
The human factors issues related to door use by people with disabilities were reviewed to identify design concerns for automated door systems. The review was limited to reports or analysis of original research, including existing literature reviews. Articles or books that only describe code requirements or design guidelines were not included.
The method used to review the literature included the following steps:
Previous literature reviews on the subject were completed by Steinfeld et.al., (1979a), Margulis (1981) and Kubasti and Steinfeld (1987). Two of these reviews covered all aspects of building accessibility, including doors. The third (Margulis, 1981) focused only on door research but included safety and security as well as accessibility research. Steinfeld, et al., covered published research through 1975. Kubasti and Steinfeld reviewed all material since the first review up to 1987. Margulis did not report the methods used for his literature search. Additional relevant research reports collected by the AEL since the last review were identified. The bibliographic search was limited to the years 1987-1992 since previous reports had been identified by the earlier reviews. A variety of sources were used, including:
The review yielded 15 publications and one informal report on the subject of accessibility and usability of doors. Some additional materials were obtained with relevant research related to door use in general. These reports are listed in the Bibliography with short abstracts. Only two of the publications (Czaja and Steinfeld, 1980; Sheredos and Lyles, 1987) and the informal report described research specifically on automatic doors although all of them include material related to the design of such doors.
2.1.2 Summary of Research
A review of the existing research is helpful to identify the knowledge base that can be applied to the development of accessibility design criteria. Research has been completed on the following issues:
In this review, we will start by describing a model of door use. This model is helpful to identify all the important human factors variables. The model is followed by a summary of findings from previous research studies.
Model of Door Use
Using a door is a common activity that most people perform almost unconsciously. When door use is analyzed in detail, however, it is a task that can become quite complex. Successful completion of door use is affected by many variables including the abilities of the user, the design of the door system, the conditions of the ambient environment and the social context.
Johnson (1981) described the door use process as follows:
While this description captures the main activities involved, it is both too narrow and too simplified. Several important aspects of task performance are omitted and others need to be elaborated to represent the door use task in a comprehensive way.
The first task, perception, should be broadened to include perceiving and understanding the method of operation, including the activities to be performed by the user and the means by which the door will open and close. For example, a user must determine whether a door has to be unlatched. Both previous experience and explicit information such as labeling and overt design cues particular to the door play roles in the process. Norman (1988) calls these cues "affordances." A good example of how perception and understanding play roles in door use is the floor mat that activates an automatic swinging door. From previous experience, we understand that stepping on the mat will open the door. The mat together with guard rails on either side are affordances. They guide the user to the correct actions.
Change of walking pattern or gait is only one example of the adjustment process in which a user engages as he or she approaches the door and prepares to negotiate the doorway. In some cases, there is no change in gait, for example, at large automated sliding doors. In fact, the goal of such doors is to eliminate the need to make an adjustment. In addition to adjusting gait, however, a door user, particularly one who uses a wheelchair or who is blind, may have to make other accommodations. These may include shifting packages to free a hand for operating a door, searching carefully with cane or hand to find the opening and/or handle, or maneuvering into a position that allows a comfortable reach to the door handle or to apply force to the door itself.
Physical exertion to open the door includes reaching to a door handle or lock and grasping and operating those devices. For automatic doors, a user may have to operate switches instead, applying pressure with the body to the door itself. Enough force must be applied to both the handle, lock and the door to overcome the resistance of both the door's weight, closing hardware, lock friction or pressure differentials between the two sides of the opening.
Once a door is open, an individual may pass directly through the doorway without impediment. For people with severe disabilities or those encumbered with packages or carrying children, this task may require some maneuvering and adjustments. In particular, doors with mechanical closers may start to close before a person passes through the doorway. The individual will then have to keep the door from closing while passing through. Obstructions in the opening, such as thresholds, can also impede the ability of a person to pass through the doorway smoothly.
If a door does not close automatically, there will be a need to adjust body posture and gait to close the door from the other side and then resume movement to one's destination. There may also be a need to reach, grasp and operate the handle, operate the lock and exert force as part of the closing sequence.
Thus, Johnson's model of door use can be revised as follows:
The ambient environment and social context play major roles in successful completion of a task. They can affect the perception and understanding of door use. What might be very simple to accomplish in daylight may be impossible at night without proper illumination, for example, finding the switch for an automatic door and recognizing how it operates. Cold weather and precipitation make ground surfaces at entrances unstable and increase the criticality of the door use task. When a user is under great time pressure, speed is more important. Margulis (1981) pointed out that the presence of an "audience" in public can have a major impact on task performance.
Thus, the task of using a door is actually quite complex. It is a tribute to the abilities of the human species that it is accomplished it so often in even unfamiliar contexts without thought. But, Norman (1988) has demonstrated that even those without limitations in ability can be confounded and frustrated in our use of doors by design. He describes an incident in which a person with no disability became temporarily trapped in a vestibule because he could not perceive how the doors leading to it operated. Door use is a critical aspect of safe egress from buildings in emergency situations. Building safety codes reflect this fact through many detailed design criteria. For people with disabilities, difficulties with door use are more pronounced and often a stressful aspect of everyday experience. Automated doors can provide many benefits to people with disabilities by making the door use task easier. But, as the analysis and model above make clear, there are many human factors issues that should be addressed in the design of automated door systems.
Clear Width
Several studies have examined the minimum clear width required to pass through a doorway with a wheelchair (Walter, 1971; Ounsworth, 1973; Brattgard, 1974; Steinfeld, et al., 1979; Czaja and Steinfeld, 1980). The clear width of the door is measured as in Fig. 1. Steinfeld, et al., (1979b), in the only American study with adults, demonstrated that a 30 in. (760 mm) clear width was satisfactory for all the wheelchair users in their sample. They recommended a clear width of 32 in. (815 mm.) to provide tolerances for fast movement.
ADAAG follows this recommendation. The recommendations of European studies ranged from 30 in. (760 mm) to 31.5 in. (800 mm.). The European Manual for an Accessible Built Environment (Wijk, 1990) has a slightly larger recommendation, 33.5 in. (850 mm.). It is not known if this was based on any recent research. An interesting twist in the European manual is the allowance of narrower doors if the doorway has a double door. The overall clear width of the double door entry can be 59 in. (1500 mm.) wide (i.e. each door approximately 30 in. clear).
Maneuvering Clearances
Several studies demonstrate that the direction of approach to a swinging (hinged) door determines the amount of space needed. The two sides of a doorway are called the pull side and the push side. On the pull side, the door is pulled toward the user to open it. On the push side, the door is pushed away from the user. Approaching the door from the pull side often requires more space than from the push side because the user has to maneuver around the door.
There are two important maneuvering clearances (see Fig. 2). The first is the front clearance, the distance from the face of a closed door to the closest obstruction opposite the door. The second is the latch clearance. This is the distance measured parallel to the closed door from the edge of the door at its latch to the closest obstruction. There are also six different approaches to a swinging door based on whether the user is coming from the side or directly in front and, if from the side, whether from the latch- or hinge-side of the doorway. Research has demonstrated that the direction of approach affects the minimum clearance required (see Brattgard, 1974 and Steinfeld, et al., 1979b).
Fig. 3 illustrates the recommendations of several studies on maneuvering clearances. Variation in recommendations can be attributed to differences in methods and research subjects (Steinfeld, et al., 1979b). In general, the recommendations of Steinfeld, et al., were adopted by the ANSI A117.1 1980) standard and have remained relatively the same in the technical criteria of subsequent standards and codes based upon ANSI, including ADAAG.
There are some differences between research recommendations and the ANSI criteria. Large latch clearances can result in significant increases in the size of floor areas, particularly in buildings with sleeping rooms (e.g. hotels, hospitals, dormitories). ANSI (1992) and ADAAG require only an 18 in. (455 mm.) latch clearance at the pull side of a door. Clearances for doors with closers are increased 6 in. (150 mm.) above the Steinfeld, et al., recommendation; and, with a front/push side approach, a door with a closer and a latch has to have a 12 in. (305 mm.) clearance even though a door without a closer or latch can have zero clearance. The need for these additional clearances has never been studied in empirical research. But, practical experience suggests that it is necessary in order for many people to get close enough to the door to operate the handle and use a good angle of approach when pushing the door open against the force of a closer.
Several studies have given attention to the design of door handles (Nichols, 1966; Woods, 1980; Johnson, 1981; Czaja and Steinfeld, 1980; Steinfeld, et al., 1986). All the studies have concluded that lever handles are preferable to knobs. Johnson studied emergency door hardware as well as standard handles. He found that the performance of people with disabilities using a paddle-type device near the latch side is comparable with people's performance using the standard panic bar across the door. Door users commonly make the mistake of pushing at the hinge side of a panic bar instead of the latch side. However, newer push pad devices overcome this problem.
Both the ADAAG and The European Manual require lever-handled or other hardware that allows an easy grip.
Several studies sought to identify the maximum resistance force for opening doors (Steinfeld, et al., 1979; Woods, 1980; Czaja and Steinfeld, 1980; Johnson, 1981; Steinfeld, 1986; Sheredos and Lyles, 1987; Bails, 1988).
The findings from the many studies that addressed this issue are quite diverse (see Table 1). The lack of consensus is due to the variety of research methods and samples used.
| Study | Study Type | Recommendations |
| Steinfeld, et al., 1979b (adults) | Static 1Simulated 2 | 8.5 lbf. (37.8 N.) - exterior, 5.0 lbf. (22.2 N.) - interior |
| Woods, 1980 (adults) | Static, Simulated | 7-13 lbf. (31.1-57.8 N.) |
| Czaja and Steinfeld, 1980 (elementary school children | Dynamic, Real | Avoid closers |
| Johnson, 1981 (adults) | Dynamic, Real | None |
| Steinfeld, et al., 1986 (adults) | dynamic, real | 8 lbf. (35.6 N.) |
| Sheredos and Lyles, 1987 (adult rehab. patients) | Dynamic, Real | < 5 lbf. (>22.2 N.), avoid closers |
| Bails, 1988 (children, 3-18 yrs) | Dynamic, Real | 3.3 lbf. (14.7 N.) |
Notes: 1. Static tests are completed in a stationary position; dynamic tests are conducted as an individual moves through a doorway.
2. Simulated studies are completed in a testing rig; real studies used actual doors.
Steinfeld, et al., (1979b) used a crude method to measure forces. But they discovered that many people in their sample could exert only minimal force. This led them to recommend that the maximum opening force should be based on the constraints of existing closer technology. Using a more sophisticated apparatus, Woods found that approximately 25 percent of his sample could not exert forces equal to or greater than 13 lbf. (57.8 N.). However, 25 percent of the wheelchair users (only 6 subjects) could not exert even 7 lbf. (31.1 N.).
Czaja and Steinfeld found that elementary school age children had great difficulty using doors with closers. Out of 15 wheelchair users, only seven were able to use doors alone at all. All of those children managed an 8.5 lbf. (37.8 N.) opening force. However, only 37 out of the 51 disabled children in the sample could complete the task of using a door with a closer even though two doors required forces lower than 8.5 lbf. (37.8 N.). Avoiding doors with closers for children with disabilities was recommended. Bails used a sophisticated device to measure forces in the use of a door with children. His findings confirmed those of Czaja and Steinfeld that children with disabilities have a very difficult time using doors with closers.
Johnson's study was related to emergency egress concerns. He used the most sophisticated apparatus and analysis techniques and tested both doors with mechanical closers and doors with air pressure differentials. Although he did not clearly document the degree of disability of the 24 people in his sample, he did report that several were very seriously disabled and did not normally travel to buildings on their own. Only three individuals (all of whom used wheelchairs) were not able to use doors with closers having a resistive force of approximately 25.9 lbf. (115 N.). Trials with air pressure differentials demonstrated abilities above 33.75 lbf. (150 N.) for the disabled subjects.
Steinfeld, et al., (1986) tested the use of doors in the field. About 30 percent of the sample could not manage doors with forces higher than 12 lbf. (53.3 N.). However, there were no doors tested that had forces between 8 lbf. (35.6 N.) and 12 lbf. (53.3 N.). Sheredos and Lyles studied use of doors in rehabilitation hospitals. The doors were typical hospital doors, much larger than those found in public buildings or housing. Of 25 total subjects (22 disabled, 19 wheelchair users) only seven (28 percent) were able to use doors with closers set at 5 lbf. (22.2 N.). This group includes three able-bodied subjects.
Margulis (1981) attributes Johnson's very divergent findings to sample selection and the "experimental set" of Johnson's research protocol. Johnson's volunteers apparently were less limited than those in other studies. Even though some may have been severely disabled, they still may have had more strength and agility than the subjects in other studies. Moreover, they treated the experimental task as a challenge and acted as they might in an emergency situation.
Aside from Johnson, all the other studies agree that people with disabilities have a great deal of trouble using doors with mechanical closers. Johnson's work suggests that force criteria for doors that are used solely in emergencies (e.g. on fire stairways) should be treated differently than those of ordinary passage doors. Fire code requirements conflict with accessibility needs. Usually, fire doors are required to have minimum closing forces greatly in excess of the maximum opening forces allowed. The 60 percent efficiency of mechanical closers means that a closing force of approximately 10 lbf. (44.4 N.) would translate into an opening force of over 16 lbf. (71.1 N.).
ADAAG and UFAS have reserved the section on opening forces for exterior doors. For interior doors, the 5 lbf. (22.2 N.) recommendation from Steinfeld, et al., (1979b) is used, as in the ANSI A117.1 (1986) Standard. The 1992 version of ANSI A117.1 deleted the opening force requirement for exterior doors. The lack of a requirement for exterior doors is based on the lack of research consensus on this topic and concerns in the building industry that the ANSI recommendations are unrealistic with respect to closing force.
ADAAG does have a requirement for maximum resistance force to stop movement of a low powered automated door: 15 lbf. (66.6 N.). This requirement is consistent with ANSI/BHMA A156.19, the consensus standard for power assist and low energy doors. ADAAG also references ANSI/BHMA
A156.10 for full powered automated doors. That standard has a maximum resistance force of 40 lbf. (180 N.) for swinging doors and 30 lbf. (133 N.) for sliding doors. The European Manual recommends a maximum of 11.25 lbf. (50 N.) for door opening force at entrances to buildings and 6.75 lbf. (30 N.) at interior doors.
Cohn (1978) measured the latch resistance force of doors in the field. The types of operators studied included knobs, levers and panic bars. He found forces with a range of 1.1 to 6.75 lbf. (5 - 30 N.).
Johnson's (1981) paraplegic and quadriplegic subjects could exert maximum forces slightly in excess of 22.5 lbf. (100 N.) on three different kinds of handles: levers, push plates and pull bars. For knobs, his subjects' performance decreased to about 20.25 lbf. (90 N.).
Czaja and Steinfeld (1980) discovered that a majority of the disabled children they studied could not depress a push button that required 3 lbf. (13.3 N.) to operate. The children could exert greater forces on large push plates and pull bars. Mean maximum comfortable forces for one hand ranged from 12 - 17 lbf. (53.3-75.6 N.).
Bails (1988) completed a series of studies on the forces that children with disabilities could exert. He recommended maximum forces as shown in Table 2.
| Item Tested | Maximum Force Range (3-18 years old) |
| Protruding push buttons | 2.3-5.2 lbf. (10-23 N.) |
| Recessed push buttons | 1.4-2.8 lbf. (6.4 - 12.3 N.) |
| Levers | 1.5-2.2 ft. lbs. (2-3 Nm.)1 |
1 measured in units of torque
Woods (1980) studied the abilities of disabled people to exert forces on several different types of handles. The 25th percentile ability of his subjects is shown in Table 3.
| Item | Total Sample | Wheelchair Users |
| Lever -- down | 13 lbf. (57.8 N.) | 7 lbf. (31.1 N.) |
| Lever -- up | 11 lbf. (48.9 N.) | 9 lbf.(40 N.) |
| Thumb latch | 10 lbf. (44.4 N.) | 6 lbf. (26.7 N.) |
| Panic hardware | 23 lbf. (102.2 N.) | 23 lbf. (102.2 N.) |
Steinfeld, et al., (1986) completed an extensive study on the forces that adults could exert on handles. They discovered that many people with hand and arm disabilities could only exert minimal forces without discomfort. The maximum comfortable forces that subjects could exert also varied with the type of grip shape used, the size of the grip shape, the mounting height and the operating direction. Based on the results of that research, Steinfeld and Mullick (1990) developed recommendations for maximum force limits in product design (see Table 4).
Table 4 Handle forces/Steinfeld and Mullick
| Item | Maximum Force |
| Finger push | 3 lbf. (13.3 N.) |
| Flat hand push | 8 lbf. (35.6 N.) |
| Pinch | 4 lbf. (17.8 N.) |
| Disc and span grip (twisting) | 3 lbf. (13.3 N.) |
| Hook grip (pull) | 11 lbf. (48.9 N.) |
| Power grip (pull and push) | 7 lbf. (31.1 N.). |
As with door opening forces, there is considerable variation in the results of studies done to date on handle actuating forces. The wide divergence of Johnson's findings from the results of the others are most likely due to differences in the instructions given to the subjects and sample selection. It is unfortunate that even less than minimal information on the characteristics of his sample are provided. Although Woods' findings demonstrate larger forces for the total sample, these can be attributed to the fact that his sample had a much smaller proportion of wheelchair users (less than 10 percent). His findings for the wheelchair group are more consistent with those of the other researchers (except Johnson). The most comprehensive study on this subject, the work of Steinfeld, et al., (1986) demonstrated results for adults that are generally consistent with the results for Bails' oldest group (14.5 - 18 years old).
The findings of Czaja and Steinfeld and Bails indicate that children with disabilities have very limited abilities to apply force to handles.
The various research findings generally fall within the range of Cohn's findings for actual handle resistance forces, but his range is quite large.
The ADAAG requires a maximum force of 5 lbf. (22.2 N.) for controls and operating mechanisms. This is roughly comparable to the recommendations of Steinfeld and Mullick. The European Manual has no requirements for this item.
Timing of Door Closing
Seaton (1979) studied door use by 802 people in naturalistic settings. He found that about 5 percent used both hands and about 5 percent used their shoulders to open the doors. These people were encumbered by packages.
Johnson (1981) reports that in studies completed by the Canadian National Research Council, Division of Building Research, a high correlation was found between the resistance of the door closer and the length of time required to pass through doors. In his study on door use by people with disabilities, Johnson observed that several wheelchair users had difficulty pushing open a door equipped with a closer and maneuvering through the doorway at the same time. The door began to close before they could pass through. He recommended the use of delayed action closers.
Czaja and Steinfeld (1980) observed that several of their subjects could open a door equipped with a closer having a delayed action feature but couldn't pass through. Only one of 25 subjects tested actually made use of the delay feature even though it was demonstrated to them. Doors have to be opened to a 70 degree position for the delayed action feature to become operable. Their subjects, who were all children, did not need to open the door that far to pass through it. Czaja and Steinfeld also reported that the closing door reduced the resistance that their subjects could overcome because the task required continual application of force rather than a short maximum force effort.
Bails measured the level of force applied to doors over time during use. Graphs from his research demonstrate radically different patterns for different types of subjects (see Fig. 4). They indicate the difficulty some of his disabled subjects had in opposing the force of the closer over the course of door use and how they adjusted. For some individuals, several trials were needed to accomplish the task. Each trial was accompanied by a repositioning to improve the "angle of attack."
Sheredos and Lyles (1987) reported that many of their subjects with low upper extremity function could not complete the task of using a door with a closer because it closed too quickly. Their subjects (adult patients of a rehabilitation facility) also could not make effective use of the door delay feature. They recommended development of a device to disengage the closer or reduce its closing force. They also recommended the development of a power operated closer that would only engage when a fire alarm activated it.
Sheredos and Lyles tested a low energy power sliding door operator with opening speeds of 1.8, 2.4 and 4.0 seconds and closing speeds of 3, 4.5 and 6 seconds. The door had a "bump force" of 13.1 lbf. (58.2 N.). Twenty five subjects with a variety of disabilities were used in the test. They also tested 11 power operated doors in a rehabilitation facility with a small group of 5 subjects (2 manual wheelchair users, 2 powered wheelchair users and one blind traveler who used a cane). These doors had timing of 1.6 - 5 seconds for opening and 3-10 seconds to close with 1 - 12 seconds delays. Total time from opening to closing for the 11 doors ranged from 8 seconds - 30 seconds. Many different types of automated doors were tested including low and high powered units and manual and automatically controlled units. No problems were reported with the timing of any of the doors tested either in the laboratory or the field.
ADAAG requires a minimum opening time of 3 seconds for low energy automated doors. ANSI/BHMA A156.19 has a table that is used to determine the timing of doors based on the size and weight of a door. The minimum opening and closing times are 3 sec. ANSI/BHMA A156.19 requires power assisted doors to meet the same requirements for closing. It also requires such doors to stop if the opening force on the door is released but allows doors to begin closing immediately.
ANSI/BHMA A156.10 requires safety zones at full powered doors. If objects or people enter these areas, sensors stop the operation of the door. The sensing devices must detect objects within certain specified ranges. The door must either stop or return to an open position when activated. The safety zone for sliding doors is the door closing path while for swinging doors, it is the area within the swing of the door. The standard has minimum opening time limits. For swinging doors, the minimum time is 1.5 seconds Closing times are governed by a table that factors in the size and weight of a door; however, the minimum closing time is 2 seconds to latch check plus an additional 1.5 seconds There are no minimum opening times for sliding doors. The minimum closing time for these doors is based on a formula. Using the formula, the minimum closing time for a 36 in. (915 mm.) door weighing less than 160 lbs (71 kg) would be 3 seconds.
The European Manual requires a "door speed" of 1.6 ft./second (.5 m./second) maximum which is about three times faster than the formula used for sliding doors in ANSI/BHMA A156.10. It also recommends that doors should not start to close before people have moved through them.
Control Locations
Sheredos and Lyles (1987) reported that poor location of manual switches for some of the automatic doors they studied caused difficulty. They do not, however, describe the actual problems reported. They tested a switch that allowed operation anywhere from 8-44 in. (20 mm. - 112 mm.) in height. Subjects had no problems using the device. No photographs or diagrams of it were included.
Many studies of reaching ability among the disabled population have been completed with adults (McCullough and Farnham, 1960; Floyd et al., 1966; Steinfeld, et al., 1979; Woods, 1980; Czaja and Steinfeld, 1980; Bails, 1983; Steinfeld, et al., 1986; Bails, 1988). The results of these studies vary based on the sample of subjects used. Some studies, such as Floyd, et al., used only paraplegics who have no limitations in reaching. Others used only children whose reaching abilities vary with age. For adults, these studies generally demonstrate that a top limit of 48 in.(1220 mm.) with no obstacles is generally satisfactory for either a side or a front reach. However, Sheredos and Lyles (1987) point out that many severely disabled people with wheelchairs have very limited ranges of reach or may not be able to use their hands at all. If controls are located low enough, wheelchairs can be used to activate switches. Other alternatives are automatic detectors, "hands free" switches (e.g., short range sensors) or standardized remote control devices.
ADAAG has requirements for general reach limits that apply to automated door controls. They are 48 in. (1220 mm.) for a front reach with no obstacle and 54 in. (1370 mm.) for a side reach with no obstacles. The European Manual requires all controls to be within an operating zone of 33.5 - 43 in. (850 - 1100 mm.). The basis for this range is not known.
Signage
No studies have been completed investigating the signage issue in depth. In all the work reviewed, the only references to signage were by Johnson (1981) and Sheredos and Lyles (1987). The former discussed the difficulty distinguishing the hinge side of an exit door with a panic bar and suggested signage might help users to know on which side of the bar to push. The latter reported that poor signage made use of automatic doors difficult for some of their subjects but did not describe any specifics. No research has been completed on tactile signage at doors for people with visual impairments. It is not known whether such signs can be located or whether they are effective in communicating their intent. There is no empirical research on talking signs at doors nor a comparison between tactile and talking signs. Since signage is of critical importance to automated door use, this constitutes a serious research gap.
Neither the ADAAG nor The European Manual have requirements for signage at automatic doors. However, the two ANSI/BHMA standards referenced by ADAAG have signage requirements. ANSI/BHMA A156.9 requires a sign to be visible from each side with instructions on how power assisted and low energy power doors should be operated. ANSI/BHMA A156.10 requires signage on all full power doors indicating that they are automatic doors. Swinging doors have to be marked with standard symbols to indicate the approach or "push" side and the inaccessible, or "pull" side. If the door is usable from both directions, it has to have a third standard marking. Sliding doors have to be marked with emergency use instructions.
Revolving Doors
Revolving manual doors have never been considered acceptable as an accessible means of egress. They are too small for wheelchair use and also dangerous for people with poor balance and those with visual impairments since another person using the door at the same time can control the operating speed. Energy conservation goals, however, have increased interest in the use of revolving doors. Moreover, they have certain advantages for security. In high security installations, unauthorized use can be more effectively controlled with a revolving door since the door cannot be held open while several people pass through. Several manufacturers now make automated revolving doors. Since automation allows the doors to be very large, they can be wheelchair accessible. Moreover, the speed can be constant and safety devices can be installed to slow a door for people who are limited in mobility or stop it if it becomes blocked.
Two reports are available on observations of automated revolving doors. A Canadian team observed the use of such a door at the Ottawa Airport in 1986. In the first phase of the study, the door was not operating properly and many problems with use were observed. There was confusion about which way the door would turn after it stopped since it backed up when it encountered an obstruction. Another problem was caused by a failure of the control that slowed the door down. The door did not slow down and consequently hit people moving too slowly, knocking them off balance. The research team noted that the speed control button was inconspicuous, non-functioning and was not identified properly. In the second phase of this study 2846 people were observed using the door. Less than .5 percent had some difficulty negotiating it. These were usually people carrying baggage. Only one of 35 elderly and disabled people who had reduced mobility had a problem using the door. Older people were observed using their hands in an attempt to slow down the door. Evidently, they did not understand that it was automated and did not notice the slow-down button.
Sheredos and Lyles (1987) observed five automated revolving doors in use in three different cities. In each city, a different group of disabled veterans was used to test each door. The groups ranged in size from two to five and all used wheelchairs. Some were paraplegics and others quadriplegics; some used manual and some powered wheelchairs. The test subjects found that the standard speed of four to five revolutions per minute was satisfactory. They found the slow speed (activated by a "handicapped" button.) of 1-2 revolutions per minute was too slow. One door automatically sped up when a user entered the detection field. This was a surprise but did not cause a problem. The special switch to slow the door was not obvious. While approaching the door, the users focused on the door itself and were not looking for details. The impact force of the door was from 30 - 73 lbf. (133.3 - 324.4 N.). However, this force was acceptable to the test subjects (wheelchair users). The emergency stop feature worked in all cases. Sheredos and Lyles suggest that this type of door needs to be better designed to accommodate visually impaired people and those who have walking limitations but do not use wheelchairs. Based on their observations and general familiarity with the needs of people with disabilities they proposed a series of detailed design recommendations for speed control, emergency stop features and warnings.
ADAAG does not allow a revolving door to be used as a means of passage at an accessible entrance. The European Manual allows the use of revolving doors at an accessible entry if they are demonstrated to be accessible. Neither document has detailed design criteria.
Although much research has been completed on door use, there is little research specifically on automated doors. The model of door use presented above can be useful to identify the research gaps.
Perception and understanding of door operation:
From existing research, we know that understanding the operation of automated doors can be a problem for people with disabilities and the elderly. We do not know how widespread the problem is since the existing research has observations from only a few subjects. New and innovative products like revolving doors seem to create the most serious difficulties. Although the existing ANSI/BHMA A156 Standards require signage on automated doors, we do not know if those provisions are adequate. Furthermore, no research has been done on how information about door operation should be conveyed to visually impaired individuals. Since people who cannot see use their hands and canes to learn about the operation of doors and devices they encounter, attention should also be given to safety for tactile exploration. The concept of "affordances" should be examined as a means to convey information without signage, for example, the use of automated techniques that gently coach the user toward the correct procedure without giving false cues.
Often doors equipped with power operators will also have high-tech security devices like card readers. There is no research on how to communicate the location and operation of such devices to people with visual impairments.
Altering gait, adjusting body posture and maneuvering within reach:
Although considerable research has investigated the need for maneuvering clearances in front of manual doors, no attention has been given to the clearances that might be necessary in front of low energy and power assisted doors. Standards and codes currently exempt the minimum clearances at all automated doors.
Reaching and grasping handles, switches or locks:
There has been considerable research on the use of handles, switches and locks by people with disabilities (see, in particular, Steinfeld, et al., 1987). This research includes specific studies of card slots, push buttons, keys and other devices that are used with locks.
Applying force to overcome resistance of handles, switches or locks:
Much research has been completed on this subject. Although some of the findings are divergent, they can be explained by differences in research methods and sample selection. Given the purpose and intent of an application, it should be possible to use the existing data base to make appropriate recommendations for maximum resistance forces.
Applying force to overcome resistance of the door :
This issue is very important because data on the abilities of people to open doors against resistance forces lead to a mandate for the use of automated doors. Research findings on this topic are divergent. However, it is possible to explain the differences and identify relevant data. The main problem is that the abilities of the more severely disabled population to resist forces of door and closers are very low. Closers are not currently designed with a level of efficiency that would allow all doors to close properly if the opening force were set at the limit manageable by people with disabilities on an everyday basis. Therefore, there is a rationale for requiring automated doors, particularly in elementary schools and rehabilitation hospitals. Only one study has been completed on emergency use of doors (Johnson, 1981). When compared to other studies, the findings indicate that under emergency conditions, people with disabilities can exert relatively high forces to overcome the resistance of door closers. Thus, there is a rationale for treating doors used only for emergency use differently. However, since the documentation of sample selection and findings in that study is not very thorough, it is hard to evaluate the recommendations for forces with respect to people with disabilities.
Passing through the doorway:
Research has demonstrated that passing through doors against the resistance of a closer is quite difficult for many people with disabilities, particularly children. The main problem seems to be that door users have to exert force to keep the door from closing while they are moving through the opening. One study (Sheredos and Lyles, 1987) reported no problems with use of automated doors by adults at rehabilitation facilities. But this study only used five research subjects. Given the results of other research on door closers, automated doors that close before an individual passes through them would pose a similar problem, perhaps even more serious since the automated doors close with more force. The best way to counteract this problem is not yet known. There are several solutions including longer hold open times, sensing devices and door re-opening devices.
Safety issues concerning use of automated revolving doors by persons who have difficulty walking are a specific case of this problem. These doors do not really "close." But, users can be bumped by the leaf behind them. Manufacturers have developed several different approaches to this problem but none have been evaluated in great depth.
Closing the door:
Automated doors almost always close automatically so this topic is not
