- read

Innovation spotlight: A walking cane that uses driverless car tech

Eric Jaffe 73

The Augmented Cane performs navigation assistance outdoors. (Photo: Andrew Brodhead)

Improving accessibility and personal mobility for people who are visually impaired is an ongoing challenge in cities. Infrastructure upgrades to sidewalks, street crossings, and transit systems are often long delayed by funding shortages or policy priorities. Promising tech advances, such as self-driving taxi fleets, remain years away and end at a drop-off point. Building interiors present obstacles of their own.

To help people who are visually impaired navigate both indoor and outdoor spaces, a team of mechanical engineers at Stanford University has designed a prototype called the Augmented Cane. True to its name, the Augmented Cane augments a traditional white cane by adding several wayfinding sensors currently being tested in self-driving vehicles, such as LIDAR, cameras, and GPS. The prototype is also low cost and open source.

In initial tests, published this month in the journal Science Robotics, the Augmented Cane fared well. Compared with a traditional white cane, the prototype increased walking speed by an average of 18 percent — and significantly improved obstacle avoidance — on indoor and outdoor wayfinding courses. While recognizing the need for more study and refinement, the engineering team concludes on a hopeful note:

Future developments in sensors and portable computation offer exciting possibilities for improving environment understanding and planning to guide participants in complicated navigation. The demonstration of these advanced navigation capabilities illustrates the potential of the Augmented Cane to improve independence for people with impaired vision.

Let’s explore the work further.

What they did

Given the challenges of navigating new routes — especially in complex and crowded urban environments — a large share of people who are blind or visually impaired avoid making independent trips or traveling unfamiliar routes. Standard white canes can help avoid obstacles but offer limited wayfinding assistance. Assistive devices using distance sensors can improve wayfinding but have been linked to significantly slower walking speeds, as they require people to process lots of information while on the move. Guide dogs are great but can be very expensive.

The Augmented Cane was designed to build on these existing navigation-assistance options at a much lower cost — just $400 for the prototype, compared to $6,000 for some electronic aids and more than $40,000 for some guide dogs. The Augmented Cane incorporates four types of sensors and provides several real-time feedback methods to improve obstacle avoidance and wayfinding, enabling people to travel at faster speeds than they can with traditional assistive devices.

Here’s how it works. As a person walks, the Augmented Cane measures distance from obstacles (via LIDAR), spatial orientation (via an inertial measurement unit), and outdoor position (via GPS); the prototype also captures images (via the camera) of important objects, such as stop signs. The Augmented Cane collects all this information through a microcontroller, processing it through wayfinding algorithms to help guide the user along a route.

The prototype’s primary form of guidance is called “grounded kinesthetic feedback” — in short, a motorized wheel at the tip of the cane applies torque to steer the user left or right. The wheel can be adjusted based on desired walking speed, and the grounded feedback can be turned off with a button. Audio feedback is also occasionally delivered via an earbud, providing simple alerts like “obstacle ahead.”

These features emerged during three iterative co-design sessions with a participant who has impaired vision — though the researchers caution the need for further development in collaboration with a large, diverse group of people who are visually impaired.

What they found

The researchers studied the prototype by recruiting participants to travel through a series of indoor and outdoor courses. Some study participants were people with visual impairments (referred to in the study as experts); others had unimpaired vision (called novices). Each participant navigated the course twice — once using an Augmented Cane, and once using a traditional white cane with standard audio feedback. The order of cane use was randomized across all the trials, as was the course configuration.

In the first indoor trial, both expert and novice participants navigated a hallway course more successfully with the Augmented Cane. On average, participants using the prototype traveled the course in less time, covered less total distance, and made less contact with environmental barriers than they did while using the white cane. Augmented Cane users also walked at a much faster speed: 38 percent faster for novices, and 22 percent for experts.

In a second indoor trial, designed with more obstacles, participants once again walked faster using the Augmented Cane than the standard cane: 20 percent faster for novices, and 10 percent for experts. The prototype also guided participants away from obstacles earlier in the route, leading to a smoother journey.

In the third trial, conducted on an outdoor course, the Augmented Cane performed even better. The prototype improved all mobility metrics compared with the standard cane, with novices traveling 46 percent faster and experts 23 percent. Participants using the Augmented Cane also made significantly less contact with environmental obstacles, with six times fewer interactions for novices and five times fewer for experts, on average.

Averaged across all trials, participants who are visually impaired walked 18 percent faster using the Augmented Cane than using the white cane — despite having years more experience with the latter.

The Augmented Cane’s grounded kinesthetic feedback method steers users through a motorized wheel. (Photo: Andrew Brodhead)

What comes next

The research team attributes the improved mobility of participants using the Augmented Cane to several factors. The grounded kinesthetic feedback, informed by the environmental sensors, made for more accurate steering assistance and less contact with obstacles, keeping participants closer to the ideal traveling route. The automated steering assistance also likely reduced the mental load involved with many electronic travel aids.

Finally, the researchers believe their decision to build on the success of a white cane, rather than starting entirely from scratch, reduced the learning curve for participants who are visually impaired.

The biggest drawback of the prototype seems to be its weight. In follow-up surveys, expert participants noted that the Augmented Cane was too heavy, which could make it harder to use for long periods of time. And of course the real world isn’t made up of short trial courses: the researchers recognize the need to evaluate the Augmented Cane “with a larger number of participants during their normal daily activities.”

Even if the Augmented Cane becomes a big hit, there’s still plenty that cities can do to improve accessibility. In addition to the infrastructure upgrades mentioned earlier, innovations like wayfinding beacons, curbless streets, and heated pavement can all provide greater mobility freedom. And not just for people who are visually impaired — the best thing about designing for accessibility is that it’s really about designing for everyone.

Follow Sidewalk Labs with our weekly newsletter or subscribe to our podcast, “City of the Future.”