Bitsensing delivers a better radar solution

Color-coded pixel cloud generated by AIR4D system (Photo: Bitsensing)

Bitsensing is a South Korean company working on next-gen radar technology which can be applied across a series of business sectors and applications, including automotive/transport, traffic management and healthcare.

The fundamental idea behind the tech is that radar as a detection tool can do a lot more than current 2D radar systems. To deliver this improved functionality, Bitsensing is working on improving sensor design, through to production-ready systems which incorporate the new capabilities.

New product launch

In May this year, the company introduced its latest AIR4D system. As the name suggests, this builds on the basic 2D data returned by standard radar systems to deliver a high-resolution 4D representation of a vehicle’s surroundings. Using point clouds, Doppler information and other sensor inputs, the tech delivers the distance, velocity and importantly, the elevation of objects in proximity to the vehicle.

Using AI to combine that data with that collected from onboard vision systems, the information can be used to build a database which can classify specific objects as they are encountered.

Jaewoo Oh, Bitsensing

“In practice, we would deliver our radar solution to a Tier 1 that would incorporate our sensors and related tech into their onboard computing stack and that would be delivered to the vehicle OEM,” explained Jaewoo Oh, senior manager for Business Development at Bitsensing who spoke with Power Progress at IAA Transportation 2026 in Hanover, Germany.

In addition to completing interrogation of outside objects, the AIR4D system can also act as an ADAS (advanced driver assistance system) tool. In much the same way as the technology can be used in healthcare to wirelessly monitor a patient, similar functionality can monitor driver condition and performance.

“I understand that some of the larger Tier 1 companies are already delivering radar solutions to their OEM customers which include ADAS functions,” says Oh.

The system, though, has primarily been developed to support vehicle autonomy. “We are strongly focused on autonomous driving,” says Oh. “To deliver that, our systems provide far more information than those now on the market.”

Point cloud data

A video on the Bitsensing stand at IAA Transportation showed a real-time recording of data points generated by the onboard system as the vehicle moved through its direct environment. Such was the detail that even with just those individual pixels it was possible to see buildings, road intersections, other vehicles and pedestrians.

Asked if the system would be capable of analysing all of this data in real time, Oh explained there was work going on behind the scenes to streamline system recognition. “We have more than 60 engineers – about 80% of our employees – working on analysis of the signals. The perception element of the system is not instant, there’s a lot of work behind that.”

Where a camera-based system will miss obstacles due to poor vision, the radar-based system from Bitsensing will recognise the problem, all at a longer range.

Oh says that the engineering team is getting some assistance from AI, but for now it’s only used in a supporting role. “We don’t use AI in our system – yet,” he says. “We’re providing information for AI learning and that will be assimilated by the integrator. The point clouds will drive AI learning; by identifying what’s wrong, we will know what’s right.

“We have a system that’s in testing now in the United States. We believe that our solution is better than others in supporting AI learning, which makes them inherently better than those already being built into vehicles today.”