A team of researchers has developed an affordable smart cane that can guide its user and identify hazards using similar technology to self-driving cars.
This is not the first smart cane, but previous models have been bulkier, weighing up to 50lb (23kg), and much more expensive.
The cane developed by the team from Stanford University can be assembled at home.
It uses open-source software and commercially available, off-the-shelf parts, including a range of cutting-edge sensors.
It weighs just 3lb (1.4kg) and comes with significantly lower costs while adding a number of improvements.
Existing smart canes can only detect objects right in front of the user, whereas the new device can not only tell you when there is a potential hazard, but can also identify it and navigate the user around it.
One of the main sensors used by the cane is LIDAR, the laser-based positioning technology already utilised by many autonomous car and aircraft systems.
It is also equipped with a range of other sensors, including the kind of gyroscope used in smartphones, GPS trackers, accelerometers and magnetometers.
AI makes navigational decisions and steers a motorised wheel
These allow the device to monitor metrics such as the user’s position, speed and directions, while artificial intelligence systems using techniques such as simultaneous localisation and mapping (SLAM) help the cane to make navigational decisions.
The tip of the cane contains a motorised wheel that can move in any direction and remains in contact with the ground.
This is able to lead the user around detected obstacles by giving small nudges or tugs to the left or right.
As well as detecting and avoiding obstacles that are in the way, the cane can use its inbuilt GPS and mapping systems to guide the user to a predetermined spot such as a store or coffee shop, in much the same way that a satnav can guide a driver to their destination.
Volunteers, including visually impaired people and sighted people wearing blindfolds, took part in a number of trials, completing tasks based on everyday navigation, such as traversing hallways and reaching waypoints set outdoors.
The walking speed of volunteers with visual impairment increased by a fifth while using the cane, with the speed of sighted participants wearing blindfolds increasing by more than 30%.
The researchers intend to make the cane entirely open-sourced, with anyone able to download the code, electronic schematics and bill of materials free of charge.
Today’s news was brought to you by TD SYNNEX – the UK’s number one solutions distributor.
