The Open Accessible Pedestrian Signals (Open APS) project seeks to address the Call for Innovation - Enhancing Mobility for Blind and Low Vision Pedestrians challenge. We provide free and open tools for generating the hardware and software assets needed to produce accessible pedestrian signals (APS) at a dramatically lower cost than the existing solution, with more features.

As of December 31, 2016, there are APS units installed at 209 intersections citywide, 78 of which were installed over the past year.

Between 2012-2016, NYC added roughly 25 accessible intersections per year. In 2016, the city committed to increasing this to 75 accessible intersections per year. While we applaud this rollout, blind and low vision people deserve a more widespread accessible solution sooner. Price has been a significant limiting factor; each intersection costs about $43,000. Learn more about NYC APS Intersections.

We demonstrate a working prototype that provides APS with greater functionality than the current implementation, at a much lower cost. Our approach leverages open city data, commodity consumer tools, and an open-source approach to enable a broader deployment of APS at city intersections, giving greater accessibility for people who are blind and low-vision.

Existing NYC APS Solution

The current APS solution is a sign box installed parallel to each crosswalk. The box emits a chirp, so it is discoverable; when pressed, it will say “wait” if the crosswalk sign is not green; and emits a recognizable sound during the duration the crosswalk is green.

The existing APS solution does not provide wayfinding guidance. There is no speech announcement of what intersection a person is at. Also, there is no support for low-vision people to more easily read street names.

APS Box installed at 7th Ave and 23rd St

Cost

The current cost per intersection averages approximately $43,060. For the 78 intersections where APS was installed in New York City, the total cost was $3,358,690.

Proposal: Open APS

We propose a more advanced set of functionality at a fraction of the price. Our devices are customized to each intersection and crosswalk, giving specific signage and audio guidance. To make this cost-effective and feasible, we will automate the creation of the signage and audio content using city public records data; and we will generate these assets free of cost.

We have built a prototype APS for the North-East corner of 7th Ave and W 23rd St, facing South-West across W 23rd St. Learn more about this intersection.

Two modes: full crosswalk guidance, or orientation-only

Our solution is built to support two modes:

The full crosswalk guidance mode is enabled when the device is connected to the crosswalk light signal (hardwired connection). We can give guidance on the light status (via LED lights, voice, and haptic feedback). While this is a much better pedestrian experience, it may incur significantly higher installation costs.

The orientation-only mode is our fall-back state. The APS gives tactile wayfinding guidance, and when pressed gives audio instructions about the location, hazards, and orientation; but we will not be able to provide guidance for when it is safe to cross. Devices may be installed in orientation-only mode in locations where it is too expensive to connect to the traffic signal; and if power goes out, the device can continue to operate off a backup battery in orientation-only mode.

Signage Features

Audio Features

Our APS emits different chirp sounds to indicate walk / don’t-walk.

The sign is a large button that speaks when pressed:

Longer presses are used to request more in-depth information:

Example long-press audio for don’t-walk:

Transcript: “Wait to cross W 23rd St at 7th Avenue, Wait the Cross Walk is 52 feet long and goes South-West”

Low Cost

Our prototype’s total cost of goods (COGs) was $66. Our open solution does not have any licensing or annual fees. Therefore, at our current COGs, building 8 APS boxes for a 4-way intersection would cost about $528.

Note that at scale, these COGs would be significantly lower; and also, the enclosure and electronics would almost certainly be mass-produced, further lowering the cost.

Assumptions

Like the existing APS installations, our solution will need power (~3 watts) and a hardwired connection to the crossing signal state (while timing can be predicted, it’s a safety issue to avoid e.g. maintenance changes or clock skew). We do not require internet connectivity; the devices work offline.

Benefits

Compared to the current NYC APS solution, our proposal has a number of benefits: