Microsoft funds AI navigation app to aid visually impaired
Microsoft has funded iMerciv’s development of an AI accessibility app designed to help visually or mobility impaired people navigate crowded city streets.
The five-year CA$25mn project is aimed to design AI tools that enable people with disabilities. The technology will offer more independence and support to those it could benefit most.
iMerciv is working on MapinHood, a navigational tool for mobile devices that allows a pedestrian to more effectively navigate city streets, aimed at offering assistance for the visually impaired or for people using wheelchairs. The MapinHood app intends to use the funding to pay for Azure machine learning licences to improve learning, storage and virtual machine capabilities along with other AI components needed in the future.
The app will use audio cues to warn users of obstacles on a path such as stairs, low hanging branches, street lamps, benches, water fountains and road signs that could otherwise cause trouble for disabled pedestrians.
“We have built a fully-flexible, customized routing engine that only caters to pedestrians. The MapinHood app is essentially a crowd-sourced mobile mapping platform built for pedestrians by pedestrians,” said Arjun Mali, iMerciv Co-founder.
Current navigation systems use satellite positioning to best ascertain a user’s location and calculate the shortest and most efficient route to reach a destination, over longer distances. Mali has seen a gap in the solutions offered: users wanting to have a route with more accessibility, cohesively mapped over short journeys, are not being catered to.
Mali plans to make iMerciv’s navigation as interactive and, as the name suggests, as immersive as possible. MapinHood is currently in the alpha stages with testing being conducted through four of iMerciv’s partners. Canadian National Institute for the Blind, Spinal Cord Injury Ontario, Walk Toronto and Microsoft have all partnered together to contribute to the success of iMerciv, with Microsoft also offering its own AI and machine learning developments.
The company is expecting to have the app in the hands of consumers by March 2020 on iOS and Android, in Toronto and the Greater Toronto area. If you’re interested in the app can visit the website and subscribe for more information.