Stanford Researchers Design a High-Tech White Cane That Nudges Users Away From Obstacles

Stanford Researchers Design a High-Tech White Cane That Nudges Users Away From Obstacles
A shot of the Augmented Cane performing navigation assistance outdoors. (Photo: Andrew Brodhead)

Engineers at Stanford University have attempted to improve the standard white cane, used to help visually impaired people move around independently and safely. In a new study this week, their design, simply named the Augmented Cane, seemed to improve the walking speed and navigation skills of both sighted and visually impaired people. The scientists say their open source design should make it easier for this technology to eventually reach the visually impaired community.

Canes for navigation have been used for centuries by some visually impaired and blind people. But it wasn’t until the early 20th century that the white cane — the white is meant to make the cane most easily noticeable to others — became synonymous with visual assistance. Despite this long history, only a small percentage of eligible people use white canes (8%, according to one estimate). Canes aren’t a perfect solution for detecting obstacles while walking, even after the extensive training needed to use them, and there are other methods for navigation, such as a guide dog, that may be preferred.

Patrick Slade, a PhD student in robotics at Stanford, had been focused on ways to improve the mobility through exoskeletons or prostheses. But during his time at Stanford’s Intelligent Systems Lab, he began to learn about new developments in improving autonomous vehicles. It made him wonder whether some of this work could be applied toward improving the classic cane and making it more usable for a wide variety of environments.

“The challenges of the visually impaired community are well documented in research literature, but their methods of navigation and preferred solutions depend on a lot of factors such as level of impaired vision, physical fitness, income, location/walkability of their neighbourhood, etc,” Slade told Gizmodo in an email.

The omni wheel allows users to choose different walking speeds and employs haptic feedback to nudge users to steer left or right. (Photo: Andrew Brodhead) The omni wheel allows users to choose different walking speeds and employs haptic feedback to nudge users to steer left or right. (Photo: Andrew Brodhead)

The Augmented Cane, which is described in a new study published in Science Robotics, has two major differences from a typical white cane. Near the base, there’s a device filled with various sensors, including a camera, that collect information about the environment around the person, including GPS and LIDAR data (LIDAR uses reflected laser light to estimate how far away other objects are). At the foot of the cane, there’s a omni-directional wheel that comes with settings to adjust for a person’s walking speed and haptic feedback to nudge the user to steer left or right as needed (the same basic concept is seen in video game controllers that vibrate in response to game actions). In theory, the device should pick up on potential obstacles ahead and assist in navigating unfamiliar places.

To test out their cane, Slade and his team had visually impaired and sighted people (24 in total) complete a series of navigation challenges in both outdoor and indoor environments. Sighted people were novices at using a cane, while those visually impaired had at least a few years of experience; all of them were blindfolded beforehand. The Augmented Cane appeared to outperform the white cane during these tests.

“Our experimental findings showed that across a range of indoor and outdoor tasks people with impaired vision chose to walk faster when using the Augmented Cane compared to a standard cane. This means our device provides some mobility benefits,” Slade said. “In addition, we did some experiments to show our device could provide assistance that a normal cane could not (for example routing to a specific room or object in an indoor setting like going to a coffee shop in a mall).”

There have been attempts to develop so-called smart canes in the past. But one major stumbling block to their widespread use has included price, with some devices costing over $1,000, compared to the standard cane, which can cost as little as $20. Some researchers have also criticised the lack of involvement from visually impaired cane users during development, which can lead to devices that don’t really provide many added benefits.

The Augmented Cane steering a participant in a hallway in front of the researchers. (Photo: Andrew Brodhead) The Augmented Cane steering a participant in a hallway in front of the researchers. (Photo: Andrew Brodhead)

Slade says that his team’s cane was co-designed with the help of visually impaired people, who provided feedback on what they would want from a smart cane. He noted that any single design is unlikely to fit everyone’s needs, but they tried to cover as many bases as possible.

“One result I found surprising was that the participants with impaired vision had an average of 20 years of experience using a white cane, but self-selected a faster walking speed with our device after only 5 minutes of training. This suggests to me that our haptic feedback method is accurate and that relaying information about the environment to the user might improve their confidence and allow them to walk faster,” Slade noted.

The team also uploaded their design to an open source database, meaning that others can now improve on it. They’ve already identified some possible avenues for upgrades, provided they can get more time and resources to develop it further. In particular, they aim to make future versions of the cane lighter.

“Although our device is relatively low cost ($US400, open-source, and lightweight (~2 lbs/0.9kg) relative to other research devices, it definitely has room for improvement to become a product. For example, every participant commented on the weight of the prototype in a follow-up survey,” he said. “I think the technology and software is available to make it a really friendly, easy to use device — it would just take some additional product development and user testing.”

In the future, he added, it’s possible that the device might only require something as simple as a smartphone and wheel attachment to work, making it easy for users to add to their existing cane and to remove when not needed.