in , ,

WaveSense develops ground-penetrating radar for vehicles, robots

WaveSense develops ground-penetrating radar for vehicles, robots

Driver-assistance systems and autonomous vehicles in development use multiple sensors for navigation, collision avoidance, and monitoring the state of the vehicle and the driver. These include cameras, lidar, and radar. Radar offers advantages in being able to see through rain and snow, but what else can it do? WaveSense Inc. is commercializing ground-penetrating radar that it says can provide additional useful data to vehicles and robots.

“We’re building the most robust and reliable positioning system for vehicles,” said Tarik Bolat, CEO of WaveSense. “It’s based on research originally done at MIT Lincoln Labs with military vehicles, and we got signals from the automotive industry that the technology should be brought into autonomous passenger vehcles.”

WaveSense looks beneath the surface

Unlike other sensing methods, ground-penetrating radar (GPR) is unaffected by snow, heavy rain, fog, or poor lane markings. In combination with GPS, cameras, and lidar, GPR can significantly reduce navigation failure rates for autonomous and driver-assist systems, In combination Somerville, Mass.-based WaveSense.

“WaveSense looks beneath the road, which is important for mapping, where you need features to be differentiated and stable over time,” Bolat told The Robot Report. “While other sensors can handle obstacle detection and avoidance, our system is set to look down for accurate positioning, even in GPS-denied environments.”

“We use ground-penetrating radar to map the surface, soil density, and utility infrastructure to a depth of 2 to 3 meters or 10 feet. That gives a more unique, stable fingerprint to track,” said Byron Stanley, co-founder and chief technology officer at WaveSense. “We have a 5cm [1.96-in.] lateral accuracy.”

“GPS isn’t reliable enough for self-driving cars,” said Stanley, who worked on GPR at MIT before WaveSense spun out in 2018. “The trend is to use something that’s physically independent — we’ve moved from inertial measurement GPS and from there to GPR with lidar.”

“We can use shrinking down data for localization to 100KB per kilometer,” he said. “The system will likely be embedded as a layer in maps that others are using.”

WaveSense is building maps of “high-value areas” and is expanding its fleet to meet partner demands, Bolat said. “We want to crowdsource map expansion,” he added. “A core set of maps would be factory-loaded, but every vehicle could help create a new map.”

WaveSense map
WaveSense is building maps with ground-penetrating radar. Source: WaveSense

Revving up radar

“The technology challenge now is that we’re at the phase where it’s the ‘automotivization’ of the product,” Bolat said. “The prototypes worked very well, but getting a system into the automotive supply chain has been a big challenge for many startups, including packaging it and getting it qualified at a component level.”

WaveSense currently has pilots with OEMs and Tier 1 automotive suppliers that it said will apply GPR to advanced driver-assistance systems (ADAS) soon.

“We’ve been working for the past two years with the big automakers on Level 2+ autonomy, which includes automated parking and active lane centering,” Bolat said. “We’re even looking at off-road applications such as agriculture.”

WaveSense roads
GPR could make localization for self-driving cars more accurate. Source: WaveSense

Other GPR applications include robots

WaveSense’s GPR could also help utility and construction companies, as well as delivery robots, said Bolat.

“With a higher-quality, closer to real-time map, it affects where to dig and the state of infrastructure or potholes,” he said. “‘Do we need to follow our maintenance schedule this year, or can we put some things off?’ We are in conversations with municipalities and utilities.”

“We’re also in talks with sidewalk delivery, mining, construction, and warehousing companies,” Bolat said. “Autonomous vehicles and robots with GPR could be more accurate in port settings. These are all examples of simpler perception situations but more challenging localization environments.”

Source: www.therobotreport.com

What do you think?

287 points
Upvote Downvote

Leave a Reply

Your email address will not be published. Required fields are marked *

This site uses Akismet to reduce spam. Learn how your comment data is processed.

Meet the Entrepreneur Who Wants to Make the World's Roads Safer

(podcast) Meet the Entrepreneur Who Wants to Make the World’s Roads Safer

Helm.ai raises $13M on its unsupervised learning approach to driverless car AI

Helm.ai raises $13M on its unsupervised learning approach to driverless car AI