While most robotic systems currently rely on optical light sensors to know where to go, a prototype robot has been built by researchers from the University of California (US) that uses wifi signals for highly accurate indoor navigation.
Light or visual-related sensors such as cameras and LiDARs struggle in low light, variable light or repetitive environments, such as long corridors or warehouses. To combat this, the researchers used wifi technology, with radio frequency signals helping to improve robotic navigation in otherwise challenging indoor environments. It’s a technology that could also be less expensive and power hungry than LiDAR.
“We are surrounded by wireless signals almost everywhere we go,” says project leader Professor Dinesh Bharadia. “The beauty of this work is that we can use these everyday signals to do indoor localisation and mapping with robots.”
“Using wifi, we have built a new kind of sensing modality that fills in the gaps left behind by today’s light-based sensors, and it can enable robots to navigate in scenarios where they currently cannot,” adds Aditya Arun, lead author of the study.
The prototype was built using off-the-shelf hardware (a possible weekend DIY project for you?), and consisted of a robot equipped with wifi sensors constructed from commercially available wifi transceivers. Using wifi access points in the vicinity, the robot was able to transmit and received wireless signals. This continuous back and forth communication is similar to a game of Marco Polo, and it’s this that makes the use of wifi signals for navigation accurate.
“This two-way communication is already happening between mobile devices like your phone and wifi access points all the time — it’s just not telling you where you are,” says Roshan Ayyalasomayajula, co-author on the study. “Our technology piggybacks on that communication to do localisation and mapping in an unknown environment.”
The team performed a test-run of the robot on the floor of an office building, placing several access points around the space, and equipping the robot with wifi sensors, camera and LiDAR for comparison. The accuracy of the wifi signals for localisation and mapping were on par with the camera and LiDAR.
“We can use wifi signals, which are essentially free, to do robust and reliable sensing in visually challenging environments,” says Arun. “Wifi sensing could potentially replace expensive LiDARs and complement other low-cost sensors such as cameras in these scenarios.”
The team is now looking to combine wifi sensors, for accuracy, with cameras, for visual and contextual information of the environment, to develop a more comprehensive, but inexpensive technology for mapping.
This research will be presented at the 2022 International Conference on Robotics and Automation (ICRA), which will take place from 23-27 May in Philadelphia (US).