You might be free to share this text beneath the Attribution four.zero Worldwide license.
A brand new approach makes use of a novel mixture of WiFi indicators and accelerometer expertise to trace gadgets in near-real time, researchers report.
The approach, which may measure velocity and distance in indoor environments, might assist enhance navigation applied sciences for robots, drones, or pedestrians looking for their means round an airport.
“We name our strategy WiFi-assisted Inertial Odometry (WIO),” says co-corresponding writer Raghav Venkatnarayan, a PhD pupil at North Carolina State College. “WIO makes use of WiFi as a velocity sensor to precisely observe how far one thing has moved. Consider it as sonar, however utilizing radio waves, quite than sound waves.”
Many gadgets, reminiscent of smartphones, incorporate expertise known as inertial measurement models (IMUs) to calculate how far a tool has moved. Nevertheless, IMUs endure from giant drift errors, which means that even minor inaccuracies can shortly change into exaggerated.
Wifi-assisted inertial odometry is a software program program, however researchers created a prototype for preliminary testing with different gadgets. (Credit score: NC State)
In out of doors environments, many gadgets use GPS to appropriate their IMUs. However this doesn’t work in indoor areas, the place GPS indicators are unreliable or nonexistent.
“We created WIO to work at the side of a tool’s IMU, correcting any errors and enhancing the accuracy of velocity and distance calculations,” says co-corresponding writer Muhammad Shahzad, an assistant professor of laptop science. “This enchancment in accuracy also needs to enhance the calculations relating to a tool’s exact location in any indoor setting the place there’s a WiFi sign.”
The researchers needed to check the WIO software program however bumped into an issue: they might not entry the WiFi community interface playing cards in off-the-shelf gadgets reminiscent of smartphones or drones. To deal with the issue, the researchers created a prototype system they might use at the side of different gadgets.
The researchers discovered that utilizing WIO improved a tool’s velocity and distance calculations dramatically. For instance, gadgets utilizing WIO calculated distance with a margin of error starting from 5.9% to 10.5%. With out WIO, the gadgets calculated distance with a margin of error from 40% to 49%.
“We envision WIO as having functions in every part from indoor navigational instruments to health monitoring to interactive gaming,” Venkatnarayan says.
“We’re at the moment working with Sony to additional enhance WIO’s accuracy, with a watch towards incorporating the software program into off-the-shelf applied sciences,” says Shahzad.
The researchers will current their paper on the work at UbiComp 2019 in London, UK.
Assist for the work got here from the Nationwide Science Basis.
Supply: NC State