ADA Requirements for Effective Communication

The Americans with Disabilities Act (ADA), its amendments, and supporting regulations set clear requirements for communicating effectively with people who have disabilities.  The law requires that communication is equally effective for people of all abilities (28 CFR §35.160).  The requirements apply to all covered entities (which include state and local governments as well as business and nonprofit organizations that serve the public) and cover any interaction with these entities, including those by riders, employees, and members of the general public.

The ADA makes it illegal to discriminate against someone on the basis of their ability.  As it applies to effective communication, the law requires an equal level of communication be made available to all persons regardless of ability.  So regardless of whether you can see and hear like most people, you are blind, or are deaf, you are entitled to get the same level of information communicated to you.

Effective communication for people with hearing impairments

Requirements apply to support those who may be deaf, are hard of hearing, or have some other sort of hearing impairment that affects their ability to hear.  28 CFR §35.160(b)(1) requires entities to make appropriate "auxiliary aids" available to ensure effective communication, including a qualified sign language interpreter, oral interpreter, cued-speech interpreter, or tactile interpreter; real-time captioning; written materials; or a printed script of a stock speech/pre-scripted announcement.

In terms of transportation facilities, this requirement most readily applies to any audio announcements that may be played throughout the facility.  Audio announcements must be made accessible to people who are deaf or otherwise hearing impaired.  This applies to all announcements at all times, including those that are informational, generic or filler in nature, or specific to a particular service, including during times where there are service interruptions or other changes to normal operations.

While the law requires entities to tailor communication based on the individual needs of the person with different abilities, the most obvious way to communicate with people who may have hearing impairments is through the use of real-time captioning, where the content of audio announcements is made available in identical text form while the announcement is being made.  This way, people who can't hear can simply read the message and get the same level of communication as those who are listening to the announcement.

Some transportation facilities are beginning to realize the importance of text messaging.  NJTransit has for many years included text transcriptions of all of its automated announcements.  This system is used at many stations across the system, including most of the busiest stations aside from New York, including Hoboken, Newark, Newark-Broad Street, Secaucus Junction, Trenton, Metropark, Atlantic City, and many other heavily used (or newly renovated) stations.

NJTransit Trenton departure and arrival screens scrolling a text version of the automated announcement playing

Amtrak goes even further as part of their new, standardized Passenger Information Display System (PIDS) being implemented at stations across the country.  The PIDS system includes a dedicated screen for text messaging paired with all departure/arrival boards and gate screens.  These messaging boards scroll the text of every announcement being played through the PIDS system.  This system is currently in use in the Amtrak portions of NY-Penn Station, as well as a host of other stations across the country big and small, including Baltimore-Penn Station, Aberdeen, MD, Denver, CO, and many more.

An Amtrak PIDS text messaging display at NY-Penn Station (Photo: The LIRR Today)

You can see a video of the system in action at the Saco, ME station:


Amtrak's PIDS system is probably one of the most robust and accessible station communication systems in the country, even if it is not the prettiest.  Announcements are read aloud by a computerized text-to-speech engine that formulates the announcements and recites them automatically, instead of relying on a human to make announcements.  This has the added benefit for all riders of making announcements consistent, easier to hear, and generally more reliable, since you don't have to deal with a human forgetting to make an announcement, not speaking clearly, or having to decipher different announcement styles from different announcers.  It's also cheaper, since you can just set it up once and be covered for most applications.

Amtrak has put a lot of thought into the automation of announcements through the PIDS system, even automating announcements for things like border crossing requirements for trains travelling into Canada, updates for delayed or canceled trains, and more.  Amtrak employees can also type any message into the system to be read aloud and displayed on the text messaging boards for any ad-hoc announcements that need to be made.

The MTA does an so-so job of communicating to people with hearing impairments...the NYCT Subway A division countdown clocks will display text versions of service change announcements, but do not convey arrival/ETA information announcements in text form.  B division countdown clocks do not display any text versions of in-station announcements.  The LIRR's Audio-Visual Paging System (AVPS), which like Amtrak's and NJT's systems consist of predominately automated or pre-recorded announcements, will display paraphrased text versions of pre-recorded announcements (i.e. all of those silly fluff announcements you hear while waiting for the train), usually delayed after the audio announcement has played, but don't display text versions of the automated train status announcements.  Metro-North falls furthest from the mark with only one- or two-line summaries of the few announcements that are made sometimes displayed on LED screens at stations.

Where the LIRR really fails, though, is at their major terminals: NY-Penn Station, Brkyln-Atlantic Terminal, and Jamaica, where the AVPS system is not used.  Instead, the LIRR uses a Solari board system and manual announcements at these major stations.  At the places where effective communication is the most important, due to the dynamic/day-to-day variations of operations at these stations, the LIRR completely fails to provide lots of information to those who can't hear audio announcements.  Manual announcements made by ushers that include stopping and connection information, announcements about service changes, and other informational announcements are made in audio form only, with no text component.

Effective communication onboard trains

While agencies are slowly coming around to seeing the importance of text messaging, requirements for effective communication apply not only to stations, but also trains and other types of transit vehicles.  While the law requires agencies to provide the same level of communication to passengers onbaord trains regardless of ability, agencies frequently fail to communicate effectively with people with hearing impairments onboard transit vehicles.

On the LIRR, for example, the M-7 fleet is equipped with an Automated Station Information (ASI) system (which recites "This stations is..., this is the train to...").  The text of announcements is displayed on LED screens throughout the car, so therefore would largely comply with the requirements for effective communication under the law.  But the announcements made as part of the ASI system are limited to announcing this station, the destination, the next station, and a watch the gap announcement...that's it.  Train crews routinely make additional announcements manually throughout the journey, including at the initial terminal, before, during, and after stops at major transfer points, approaching stations with short platforms, and during delays.  Here the LIRR fails badly to communicate effectively with people of all abilities.  All of these announcements often made manually by conductors (some of which, like announcements about what car you need to be in to exit at a particular station, or updates during delays or other service disruptions, are extremely critical) are never translated into text for people who can't hear.  That means they often miss out on the content of those messages.

And the extremely limited information conveyed by the M-7 ASI system only works if it's actually turned on, which is far from universal.  In lots of cases, the system either malfunctions or conductors just don't bother to turn it on/set it up properly, so the screens stay off for the entire trip.  More frequently, the automated system does not come on until only after the train has left its initial terminal.  Train crews almost never update the ASI system to reflect/announced added stops, changes in terminal, or other route changes--the system just uses incorrect information or is shut off completely.  On the LIRR's C-3 diesel car fleet, the destination sign system almost never works, meaning no information is communicated to people who can't hear audio announcements, and the railroad has been very slow to bring to fruition a plan to restore the functionality of the C-3 destination signs.

The law grants few exceptions to effective communication requirements.  In order to communicate effectively, practically all audio announcements must have a text version made available simultaneously.  You can reach the same conclusion as you can for station communication that the easiest and most efficient way to achieve effective communication is to automate as many announcements as possible.

Almost all announcements made onboard trains are predictable to the point they could be automated fully if the railroad put in the effort to achieve effective communication.  Connections generally stay the same, or could be reasonably determined/derived automatically as part of the LIRR's schedule/TrainTime system.  Announcements about short platforms definitely don't change...it's not like platforms get randomly longer or shorter overnight, and the length of trains is tracked as part of the LIRR's TIMACS system, so announcements could be tailored to the particular train.  During service disruptions, a detailed list of informational announcements could be prepared to keep people up to date, or text-to-speech announcements could be broadcast directly to trains from a central location to make sure everyone is getting the most up-to-date and accurate information about what's going on (the M-7 fleet already has this functionality, though it is not used).

Audio-only announcements should only be made in emergencies where there's an immediate threat to life.  Otherwise, there's really no excuse for not having automated announcements/real-time captioning available onboard trains.

Fully automated announcements also has the added bonus of bringing much-needed consistency and reliability to passenger communication.  The current system of leaving announcements up to the individual train crew means far too many times passengers are given incorrect information, or no information at all.  If the LIRR employs 1,300 conductors then there are at least 1,300 different styles conductors have to make announcements, since I don't think I've ever heard two different conductors make announcements in the same form.  Inconsistency begets confusion.  I can't count the number of times I have been on a train where there were zero announcements about needing to be in particular cars to get off at stations with short platforms until the train is either pulling into the station or already stopped, if there were any announcements made at all.  That means people have to frantically run between cars once they realize the door they're standing in front of is not going to open.  I also can't tell you the last time I heard conductors make announcement about quiet cars two trips in a row.  Full automation also means that informational announcements can be streamlined and standardized so passengers hear the exactly the information they need to hear and no other fluff or unimportant announcements.

Automating announcements

The act of actually automating announcements isn't all that difficult.  It requires a lot of upfront effort to think through all the different scenarios which audio announcements are necessary and developing the logic of when they should be played, but it's nothing close to rocket science.  The fact that other railroads like Amtrak and NJTransit have been able to figure it out shows it's possible.  Going the extra step and integrating schedule and train status systems into the announcement logic can make sure passengers always receive the most accurate and up-to-date information.

While it would be possible to sit somebody down at a microphone and create recordings of every possible announcement, it's really not necessary to even do that.  These days, text-to-speech engines have come so far that they bear an increasingly closer resemblance to human speech.  Amtrak and NJTransit both use text-to-speech engines.  These services have also become extremely affordable--Google Cloud text-to-speech prices speech synthesis at $0.000004 per character for typical computerized voices and $0.000016 per character for the high-qualify human-like voices.  Amazon Polly is competitively priced at $0.000004 per character, and there are even more options that offer similar pricing.  The full list of LIRR stations is about 1,333 characters long, meaning to recite the entire list of LIRR stations using Google or Amazon's text-to-speech engines would cost half a cent.  And these services will let you generate audio files once, store them, then replay them an unlimited number of times for that low rate (the LIRR could therefore pregenerate almost all of their announcements, paying just once, then use the live text-to-speech engine on the fly during unplanned changes, keeping costs extremely low).

The affordability of modern text-to-speech engines means fully automating announcements would not create an undue financial or administrative burden, meaning the railroad has little ground to ignore requirements for effective communication.

Effective communication for people with vision impairments

While effective communication with people who have trouble hearing is by far the most deficient area for most systems, it is also important to think of those who can hear but can't see.  The same requirements for effective communication extend to people who are blind or otherwise visually impaired.

While effective communication for people with hearing impairments means that every audio announcement must have an identical visual/text component, the inverse is true here: all text or imagery must be accessible to people who can't see.  For permanent signage, this could be achieved through braille (on a nearby wall for overhead signage) and raised lettering/symbology (for all signage on walls or within touching distance).  For video or informational screens where information changes, text-to-speech speech synthesis, where onscreen text or images are read aloud.

Verbalizing information on screens is a major deficiency when it comes to making information accessible to those who can't see.  While the LIRR has gotten very good at station communication, with the Audio-Visual Paging System in use at every station but major terminals, the LIRR does not do a good job of making the information shown on these displays, the Solari screens at terminals, and service status message boards scattered throughout the system accessible to those who can't hear.

The most efficient and easiest way to achieve effective communication with screens and message boards is to provide a way for people to hear a recitation of what's shown on the screen at that moment. While you could have an audio voice constantly reciting everything that's on the screen (this is what Metra does at its terminals), the most practical solution to this is to have a button near every screen that says "Push for Audio", and then then there is a voice that recites everything on the board at that time is the most practical.  This is what NYC DOT does for the new real-time bus arrival information signs popping up at some bus stops across the city (see the photo at the right).  This is the approach CTA takes in Chicago, where there is a button you can press to get the information read out to you.  An example of that system works is shown in the video below:



Adding one of these buttons with a speaker to departure boards and near every AVPS screen is an easy way to communicate this information to people who can't hear.  It's important to note that merely having roughly the same information announced occasionally does not satisfy the requirements for effective communication.  Just because the AVPS occasionally announces train status information does not mean the railroad is communicating equally.  Not only do the AVPS/usher announcements not reflect all content on the screens, but you also have to wait around to hear the information when the announcement does play (if at all).  If people who can see can get this status information on demand, so should people who can't see, which is what makes having the buttons and audio recitations necessary, even if there are also audio announcements for other things.

(Photo: DanTD/Wikimedia Commons)

Accessible wayfinding is also a major issue.  While the LIRR has small, blue braille placards at stations showing the station name (like the one pictured at the right), there is typically only one per station (even for really large stations) and they are not in consistent locations, so can be difficult to find.  And this only tells you what station you're at, not what platform you're on, where to actually board the train, where to find the exit or connecting services, etc.

Almost all wayfaring at LIRR stations does not have raised lettering or braille versions available nearby, meaning people can be at a loss for how to get to their train unassisted.  This can be especially complicated for stations with multiple platforms.  And even at stations with one platform, you still may need to find where there is a ramp or elevator to help you get up to the platform or down to the street with ease.

(Photo: The LIRR Today)

28 CFR §35.163(a) requires that information about the location of accessible services and facilities be made available in a format that is accessible to people who are blind or have low vision.  This means, at a bare minimum, information on where to find ramps, elevators, and other relevant services be made accessible.  In 2018, the LIRR launched a "LIRR Care" program to better help people with disabilities navigate through the LIRR system, particularly major terminals like NY-Penn Station.  The railroad put up a whole bunch of additional signs around NY-Penn Station showing people where they need to go to take advantage of the service, however none of those signs feature raised lettering or braille text, so people who are blind have no way of knowing where accessible services are located, or even where they need to go to find help.

Other technology, like using Bluetooth beacons, can also be employed to help people navigate facilities, especially ones that are particularly large and complex.  A system using Bluetooth beacons to help guide people with visual impairments is currently in use at London's Euston station, for example.  Adding these features into Amtrak's Find Your Way application for NY-Penn Station and repeating something similar at other complicated stations would be a great step forward to making these facilities more accessible.