The C-Walker is a high-tech walking device that aims to safely guide people with cognitive impairments through public spaces like airports and shopping centers, reducing their reliance on visual signboards and avoiding obstacles in their way. Using onboard sensors, this "cognitive navigation prosthesis" monitors its environment in real time to figure out a path that poses little risk, actively re-planning it when it encounters problems like wet floors, or people dashing about. Aside from aiding senior citizens, the technology is expected to come in handy in factory settings, helping workers avoid danger zones and accidental collisions with machines.
"As you know, GPS does not work indoors," Josef Alois Birchbauer, a researcher at Siemens Corporate Technology (which is a collaborator in the EU-funded DALi project) tells Gizmag. "And many other indoor localization systems rely on proprietary installations like Wi-Fi fingerprinting and Bluetooth whatever, which aren't really standardized. We want to empower older adults to perform tasks of daily living outside their home."
To create a smart walker that could function reliably in unstructured environments even in the absence of other technologies, the team had to come up with a cognitive navigation system that relied solely on sensors integrated into the device. When an elderly person enters a mall, for instance, they can point to a sequence of destination points, on a tablet mounted on the walker.
The walker interfaces with the private cloud in the environment and surveillance cameras to get a 3D map of its surroundings and up-to-date information on any potential issues. If there is no cloud to connect to, it takes a series of pictures with an onboard camera and keeps on track by estimating the camera's motion relative to the scene.
Using a combination of sensors including a Kinect sensor, the walker computes the most optimal route to the user's chosen destination. Keeping an eye out for warning signs, it tracks the position, speed and direction of people moving within its range and identifies obstacles. It refreshes its internal model of the environment and the people moving around in it every 500/300 milliseconds (about two to three times a second) by taking snapshots of its surroundings. If there are any anomalies such as items falling down from a shelf, a wet floor or an unusually crowded area, the walker re-plans the route.
"Navigation has two aspects, a short term aspect, that tries to avoid any obstacles, static ones, and most important moving persons [...] as well as a long term planner that is quite similar to what you actually would think of in terms of a car navigation, that basically defines the overall route to go" Birchbauer tells us.
It encourages the user to stay on the plotted route through a mechatronic guidance system that uses electro-actuated brakes. The researchers are also investigating using sounds delivered via headphones and haptic bracelets as alternatives. Test runs with cognitively intact people on the verge of a possible decline are underway. The team aims to evaluate whether enabling them to remain active with the aid of the walker will slow down their decline.
The biggest challenge, the team says, is packing in the computational resources required to carry out all the complex processing in real time. They plan to use low-cost hardware to help keep the cost of the C-Walker low.
Siemens also plans to utilize the technology in factories and industries to create a "cognitively intelligent industrial environment" that increases work efficiency and helps people and machines interact with each other in a more comfortable manner.
"Think for instance on a forklift truck that moves autonomously in an industrial environment," explains Birchbauer. "It needs localization to perform its task which is usually done with a pre-programmed control logic. With the DALi technology, the forklift truck could dynamically react to the environment properly, e.g. reroute across workers or other obstacles. Also, think of human workers that are only allowed to operate a machine or device when outside a pre-defined danger zone. If a service technician needs to inspect a large machine or device, it's helpful to navigate him to the desired position and show context-relevant information."
The multi-disciplinary team working on C-Walker includes researchers from the University of Trento, University of Northumbria, University of Siena, INRIA (Inventors for the digital world), and FORTH (Foundation for Research and Technology) working with partners from Siemens, Visual Tools and INDRA.
Check out a video of the C-Walker in action.