AiThority Interview Series with Kobi Marenko, CEO and Co-Founder at Arbe

In order for autonomous vehicles to move beyond proof-of-concept to mass production, there is no question that an advanced, dependable, and economically affordable sensor solution is required.

Tell us about your interaction with AI and other intelligent technologies that you work with, in your daily life.

It’s actually hard to answer this question from a consumer’s point of view. AI is rapidly evolving in sophistication to the degree it is becoming seamless. I know there’s AI in many services I use often — such as Amazon, Netflix, Airbnb. Think of YouTube, which is responsible for approximately 40% of mobile internet traffic. AI is responsible for 70% of their views. This alone means AI is responsible for over 25% of online traffic.

But I don’t feel it. That’s how good it’s gotten.

When did you start in the AI/autonomous vehicles space? How did you decide to co-start Arbe? 

My co-founders Dr. Noam Arkind, Oz Fixman, and I worked together in my previous company. Following its successful exit, Noam’s vision of a radar-based sensor for autonomous machines sparked the founding of Arbe in 2015. We originally focused on building radar sensors for the robotics and drone industries. A year later, we pivoted to the automotive industry. We felt it presented a larger and more compelling opportunity for us.

Originally, we planned, like everyone else, to buy radar chips. But we quickly realized that there was no chipset in the market robust enough to handle the advanced capabilities we envisioned. So, we developed our own. The chipset development efforts were accompanied by the algorithm and AI development from day one. And now, that’s making all the difference.

Tell us about Arbe’s product. How is the company’s technology evolving the autonomous vehicle landscape?

OEMs are ramping up Level 2 and 3 autonomous vehicle production, a move that requires safety-critical functions transfer from the driver to the vehicle. This cannot happen without a sensor that can instantaneously respond to the full range of driving scenarios including the ability to identify and assess risk and execute path planning while offering a non-irritating driving experience for both the driver and those sharing the road.

This has left the industry with two main problems: First, we haven’t had a sensor suite capable of that level of performance. Second, many sensor suites rely on LiDAR, which is quite expensive, at least 10x more expensive than radar. So, the price point would limit Advanced Driver Assistance Systems (ADAS) availability to premium and luxury vehicles.

What we’ve been able to do at Arbe is solve both of these problems. We’ve managed to produce an affordable sensor robust enough for ADAS and autonomous driving.

Arbe’s ultra high-resolution 4D imaging radar with long range and wide field of view sensing, along with AI and SLAM (Simultaneous Localization and Mapping) algorithms, promises to disrupt radar technology and transform the role of radar from a supportive sensor to the backbone of the autonomous sensor suite, it is a game changer.

How do you differentiate Arbe from other similar providers?

Arbe achieved a breakthrough in radar performance. We are delivering an image 100 times more detailed via higher resolution sensing. Our technology can identify and track objects small to large, moving and stationary. For example, we can identify pedestrians, bikes, and motorcycles, and separate them from vehicles and environment objects — even when they are somewhat concealed by them.

One of the most significant obstacles to achieving ultra high-resolution radar has been the amount of processing power required for the analysis of enormous amounts of information.

Our proprietary chipset and AI have made these obstacles obsolete.

Added to this, Arbe proprietary, AI-based, post-processing software stack includes a radar-based SLAM solution powered by AI. Our SLAM algorithms perform real-time clustering, tracking, and self-localization as well as false-target filtering and radar-based and radar-camera based object classification. It is important to note that performing SLAM while utilizing radar data is a patent pending innovation.

How do you see the raging trend of including ‘AI’ in everything impacting businesses?

Basically, AI is an enabler of advanced solutions that in the past required too much computational power or storage.

While AI is being applied to almost every industry, we actually consider autonomous driving one of the toughest AI challenges. There are many variables that need to be taken into consideration in real-time. For Arbe, AI amplifies our breakthrough radar technology.

What are the biggest challenges and opportunities for AI companies in dealing with rising technology prices?

For Arbe, AI price means internal resources rather than direct costs. Therefore, our challenge is to keep resource requirements modest while not compromising on performance.

In your honest opinion, how safe are autonomous cars, currently? Do you think all future cars will be designed as autonomous/driverless? 

Today we’re seeing vehicles equipped with various levels of autonomous-driving systems. Some are more advanced than others. But even when looking at the most advanced systems, none currently can replace the human driver outright. Recent AAA testing shows that the leading ADAS systems on the market fail to handle real-world driving scenarios, including those that are relatively common. Vehicles drifted out of lanes, struggled with moderate traffic, curved roads, and busy intersections.

An alarmingly 75% failed to avoid accidents when vehicles ahead changed lanes. Google’s driverless car Waymo, for example, reports that a human driver is required to get involved once every 11,000 miles. Just to put this in perspective, statistically, accidents caused by human drivers occur once every 500,000 miles. So, there’s a lot of room for improvement before we can say driverless cars are safer.

In order for autonomous vehicles to move beyond proof-of-concept to mass production, there is no question that an advanced, dependable, and economically affordable sensor solution is required. I believe that Arbe is this technological solution — the missing link — if you will. I estimate it will be at least 10 years before we see roll-outs of private autonomous vehicles.

How do you consume information on AI/ML and related topics?

I stay up-to-date mostly via Medium blog posts.

Where do you see AI/Machine Learning and other smart technologies heading beyond 2025? 

In our domain, I believe AI will play a major role in the fusion of sensor data. More generally, I expect to see an evolution towards general AI, where the code is writing itself, and machines teach themselves how to resolve problems. The end goal is a lot further than in 2025, but I would like to see meaningful progress happening in this time frame.

What is your opinion on “Weaponization of AI”?

The classic line “With great power comes great responsibility” applies. Every innovation can be used for harm.

Social media and the internet are great examples. Because of free and instantaneous communication and information sharing, we’ve been able to achieve amazing things as a society. But there is a dark side as well, which we are struggling with now in our culture and our legal systems. AI is no different. I truly hope our society will show constraint and responsibility as we advance AI and further incorporate it into our daily lives.

What AI start-ups and labs are you keenly following?

I’ve got my eye on Hailo, a Deep Learning processor company that empowers intelligent devices with the performance of a data center-class computing.

What’s your smartest work-related shortcut or productivity hack?

There’s no doubt, my work is super hectic. So, over the years I’ve developed this tactic to clear my mind of work-related tasks, this enables me to jump back in with clarity and fresh perspective. Every week I try to take a break from work to read, practice yoga, or go to an exhibition. After this, I sort of start over, able to approach my work from a new angle.

Tag the one person in the industry whose answers to these questions you would love to read.

Eyal Gura, CEO at Zebra Medical

Thank you, Kobi! That was fun and hope to see you back on AiThority soon.