Intel and GoodMaps collaborate to make indoor mapping technology accessible

This article is a feature for our dispATches newsletter - a monthly round up of assistive technology news and events​. Keep up to date by signing up to our dispATches newsletters

A recent collaboration brings the cloud computing and AI smarts of Intel together with the indoor mapping expertise of GoodMaps to help overcome the challenges that people with disabilities face when trying to navigate complex or unfamiliar indoor environments.

It's always been difficult for someone who is low vision or blind, for example, to find their way around an unfamiliar and often bewildering building. It is similarly so for people with a learning disability trying to interpret signage that almost everyone also finds confusing. Wheelchair users need to know where to find lifts and accessible bathroom facilities. Successfully navigating indoor spaces can be tough if you have a disability.

Technology has long been leveraged to help. GPS disappears inside a building - and, even in the best outdoor conditions - doesn't afford you the accuracy needed for successful indoor wayfinding. Bluetooth beacons can aid orientation and provide additional information about nearby points of interest (such as where to locate the lift's control panel; "To the left of each lift at chest height") and, given enough for successful triangulation by your phone, sufficient accuracy to help you find that reception desk, doorway or staircase eyes-free. The down-side is that they require significant initial installation (you need many dozen on any given floor to be useful) and ongoing maintenance as their batteries only last so long.

A far more efficient approach is to create an accurate digital map of an indoor space and then leverage powerful AI to help interpret that space real-time. For a user holding a device - equipped with a powerful processor, camera and lazer-finding (lidar) capabilities - the upshot is access to as much information about objects, obstacles and orientation as they personally need. The good news is that all of these smarts exist in many of the smartphones we already have in our hands.

How can technology help? 

Discover our Don't Disable Me training course series that focuses on the lived experiences of people with disabilities including those who face hearing, visual, and physical barriers. In the course, you can learn first hand how technology can support those facing these barriers at work, in study and day-to-day life.

A GoodMaps and Intel partnership

Digital mapping experts; GoodMaps and chip and AI giants; Intel are combining their strengths to help make this approach to indoor navigation a reality. Initially Focusing on students at Portland State University, GoodMaps are working to create the next generation of indoor high-resolution 3D maps that will “Help solve complex tasks in dynamic environments through AI and cloud-computing capabilities” and Intel is investing, researching, and developing the key hardware and software enablers to help interpret the environment and give each user a personalised experience based upon their needs.

For more information on the project, check out a recent article on Disability Insider.

Give me a Clew to help me navigate

As a blind person I can totally relate to the challenges of indoor wayfinding – and the use of AI to assist me with this challenge is nothing new.

For several years now I’ve been using a brilliant app called Clew that uses your smartphone’s camera combined with the smarts of AI to map and retrace any indoor route.

The shortcoming to this approach, however, is that it only works for specific routes that have been walked before. With some assistance, you could walk several routes in buildings you need to navigate and save them in the app, but if you wanted to go off-piste (to a lecture in a room you’ve not visited before, say) then you’re stuck.

This is where the power of the above collaboration comes in. Take the limited data captured by walking a route and expand it to the entire building – every floor, nook and stairwell. Then take the AI required to recognise the objects on a single walked route (seen only from those specific directions) and expand it to be able to interpret every object within a building from every possible angle and orientation. Then consider the user’s specific needs when it comes to the type, frequency and level of information they require And, voila! You have what it takes to crack the challenge of effective indoor wayfinding for everyone.

So best of luck GoodMaps and Intel and here’s to never getting lost again.

Further resources 

AbilityNet provides a range of free services to help disabled people and older people. If you can afford it, please donate to help us support older and disabled people through technology.