Sharpen Your ADAS Radar Vision

The men and women that designed the radar scene emulation technology with the miniaturized radar target simulators (mRTS) say each one of these mRTS is like a pixel in a TV screen. In similar fashion, they’re assembled into a larger image—and that’s what the scene emulator is designed to do. So, each element is a pixel, or a radar pixel. Or maybe it’s a “rixel"?

To introduce revolutionary technology, we need revolutionary words.

The Radar Scene Emulator is like virtual reality goggles for radar sensors and cars! Pretty cool huh? The excitement within the research team in Keysight is just very palpable.

They were energized by the challenge posed in developing this solution, and now are super excited about the possibility the solution will enable, its ability to accelerate autonomous vehicle development, and its benefit to society. I look forward to that day in great anticipation.

Autonomous driving is a future capability that can offer enormous quality-of-life improvements for those who cannot drive themselves. My father was diagnosed with multiple sclerosis and it has been difficult to watch the man who helped me with my math homework and taught me how to play softball, struggle as his physical ability declines over the years. I know this loss of independence is difficult for him sometimes.

Autonomous driving would help people like my father hold on to their independence, while still being safe on the road.

What really keeps me awake at night is troubleshooting through the design and test phases, and ensuring we are able to procure the necessary materials for building the system on time. Despite all the challenges, I am very excited to work on this project because our Keysight Radar Scene Emulator is a first-to-market solution.

I am also inspired by the fact that there are many opportunities in autonomous drive emulation for us to materialize our vision to connect and secure the world. We are creating a new future by supporting our automotive customers to realize their vision, and our vision of mobility.

Time flies when you're having fun, learning, and being creative. The research is ongoing, but we've already covered a lot of ground and are starting to form a much better picture of how the whole thing will come together.

Watching one or two pixels emulate a scene is something my brain can still follow and match the interaction of one or two objects. But seeing the environment with hundreds of objects is something else altogether. It's a cluttered world and it's going to take a lot of innovation and many disciplines to make sense of it all.

It's been exciting working on this project because the science behind radar and sensor fusion is fascinating. It's also fulfilling to be part of a solution that will make autonomous driving safer in the future. And this solution is also more environmentally friendly than having to do hundreds of miles of drive tests on the road just to test and verify the radar's algorithms.

The complexity of the Radar Scene Emulator introduces a lot of challenges related to calibration and software and hardware integration, but it's great to be working on those challenges on a team, and for a company, that has the expertise to solve these challenges.

I live in Silicon Valley and am used to seeing lots of self-driving startups. Their cars are on the roads driving many miles around the same residential streets with monitor drivers seating in the passenger seat. It is really not sufficient to bring the level of quality needed to replace a human driver for corner-case driving situations. And that's where our new Radar Scene Emulator is helping to make self-driving cars a reality.

The end goal for me in this solution is to help make self-driving systems sufficiently safe, so that I can think about putting a grandmother, or my relatives who really shouldn't be driving, in a driverless taxi and know they will be safe.

Full disclosure – I joined the team after the Radar Scene Emulator launch campaign started. Of course I had heard internal rumors about a 'big secret project;' some brand new solution based on awesome technology developed by Keysight Labs. I was told it was going to be a paradigm shift in how automakers test the logic in self-driving vehicles!

At first, I was enticed by the novelty of the technology behind the Radar Scene Emulator solution. Then I met the folks working on it – all of them inspiring people, from different organizations, different regions, and different cultures, working from the office, or the lab, or from home, around the clock. Each one was invested in the success of this project and passionate about the result, as am I!

#Keysight #GreatPlaceToWork #TeamWork

The artificial intelligence (AI) community tells us that autonomous driving will be safer than human driving. We'll never know if this is so, until the car companies begin to test the dangerous, and what I would call the verge-of-accident type of scenarios. But of course, the latter are totally unethical to test in reality. I've always thought that here is where electronic emulation can help.

It is my hope that our Radar Scene Emulator can present these sorts of scenarios, and others, to the radar, as the radar would see them in reality. Then, it is up to the car's AI to learn to react to the sensor information in the safest manner possible, on a case-to-case basis. Only then will the general public and for that matter, the car companies themselves, gain confidence in automated drive technology.

The Radar Scene Emulator project has been very challenging, but the best things in life often are. This project required invention in many areas, and was full of risk and uncertainties. In particular, we needed to make sure our solution emulated the real world as accurately as possible and without creating a false radar detection. Given the complex and sophisticated waveforms employed by today's radars, that was a real challenge.

After a tremendous amount of careful and detailed design, we're now seeing good results on a variety of radars. My goal for all this work is to see our technology widely adapted in the autonomous vehicle industry and to know that it's helping the engineers who are making autonomous vehicles a reality.

From the very beginning, when we started considering innovative ideas for testing autonomous drive systems, we discussed challenges around getting radar sensors into the game. What I have personally enjoyed most on this project is the interaction with lead customers and partners, which has helped shape our vision into the Radar Scene Emulator product that is now a reality.

In addition, working with a team across the globe – between the U.S., China, Malaysia, Germany, France, and Korea, to name a few – and with bright and innovative engineers in Keysight, with our partners, and customers, has made this journey so interesting. Seeing this product come together now is a tremendous achievement!