TacTile

An app paired with a special white cane tip that reads RFID-enabled pavement, providing precise navigational information for an independent wayfinding experience throughout campuses.

Get On Board

Onboarding that adapts to you

Chroma (our AI Voice User Interface) learns your name, vision level, text size, color preference, and voice pace, then walks you through pairing your cane tip.

Adjusting speech rate in the onboarding flow

Big bold colors guide the user to pair their cane with the app

Search

Search and wayfind

Find a destination and follow step-by-step, clock-referenced directions with alerts for staircases, traffic lights, dead ends, and intersections.

Primary search page

Route details

Navigation view

Sonification page with elevation, grade, and ground texture information. As you run your finger across the terrain graph, it plays a tone to indicate elevation and puts gaps in the tone based on the roughness of terrain.

Explore

Explore freely

Move without a set destination and still hear about the world around you as you go.

Explore screen

Color Customization

You can see clearly now

High-contrast black and yellow, adjustable type, and color-vision modes (Deutan/Protan, Tritanopia, and custom). Type is set in Atkinson Hyperlegible and Inter.

Display and color customization

We found most users had strong preferences for fore- and background colors, so we set those during onboarding.

12345

1. 1

   RFID sensor

   Reads the UUID embedded in each pavement tile.

2. 2

   Removable rubber ball tip

   Easily replaceable; consumable

3. 3

   Smart Tiles

   Always on, no batteries needed. Marginally more expensive than existing tiles.

4. 4

   Fingerprint sensor

   Ensures route and navigation information remains private

5. 5

   Speaker + Mic

   Audio input and output in the event the user doesn't have headphones

Let's define blindness first

Blindness exists on a spectrum of impairment. There are two types: the top row shows definition impairments, while the bottom row shows obstruction impairments. Our solution should work for all of them.

Our first big clue

The white cane helps people familiarize themselves with an environment through landmarks. But it struggles to detect water or fabric on the floor, can't warn of body-level or head-level obstacles, and makes hazardous terrain hard to avoid. This could be an interesting area to improve.

Sonification

We needed our product to work without any visual output. We found a technique called "sonification" (mapping data to pitch and timing) that allowed us to do just that. For delivery, speakers broadcast private information out loud and headphones cover the ears blind users rely on, so we landed on bone-conduction headphones that vibrate the skull to give auditory stimulus and leave the ears open.

Sensory Cue workshop

We "blinded" sighted participants with goggles and observed them as they tried to identify common household objects without using vision.

We mapped dozens of concepts on a feasibility-versus-usefulness matrix, from braille printers and guide drones to shoe sensors and RFID-tagged pavement, before narrowing to the ideas with the most promise.

Every concept plotted by feasibility and usefulness.

Our final two concepts

We developed these concepts to get higher fidelity solution feedback

Seecon — indoor beacons that notice when furniture or obstacles move, working with smart-home platforms to ding when something's out of place as you enter a room.

TacTile — RFID tags embedded in tactile paving, read by a sensor on the user's cane, feeding navigation and spatial information back to the user.

We chose TacTile: replacing tactile paving with RFID that the white cane reads, giving the user navigation and spatial information along their route.

I built the design system using bright, vibrant colors, solid iconography, and the Atkinson Hyperlegible font.

I made the interaction model

We made a tone of voice map for the VUI. Interestingly, there are some times to be more dominant (like when giving navigation instructions)

Our VUI's prosody (word choice and speaking pace) map

1234

1. 1

   Positive

   Good type size

2. 2

   Negative

   Needs pagination indicator

3. 3

   Negative

   Needs some kind of global navigation

4. 4

   Negative

   Unclear how to get out of home page

12

1. 1

   Negative

   Walking needs to be prioritized

2. 2

   Positive

   Icons are clear

12

1. 1

   Positive

   Route line is clear and easily understood

2. 2

   Negative

   Doesn't communicate "emergency" well

Cane Improvements

Our original idea was an omnidirectional-wheeled 'Omnicane' with a sensor to talk to the RFID pavement.

We moved to a standard white cane with a replaceable rubber ball tip that holds the RFID sensor and added mic input, audio output, and a fingerprint sensor.

Round 1 told us users wanted simple, streamlined information with the option of landmarks or general directions, and more precision throughout. When things went awry, they wanted a panic button to contact help, and restroom navigation was a surprisingly big pain point.

After a few more rounds, we were able to develop a significantly more robust experience, and even had an expert look over our concept.

We started designing with the blur filter on so we could simulate low vision

Lots of testing

We did cane-based testing by blinding sighted people and having them walk through an unfamiliar environment.