Rahul Dutta
Simple, Curious and occasionally Insightful

Israeli technologies are making self-driving cars safer.

Self-driving cars will soon become a daily part of our lives

It is almost surreal to imagine hopping on to a car that drives itself to your work, while you are playing your favourite game or reading your daily quota of morning news. This scenario may not be too far from becoming reality if the buzz around self-driving is true.

The idea of breezy commutes where autonomous vehicles will make road accidents a thing of the past is garnering a lot of public interest. The promise of no crash, lowered emission, and reduced traffic congestion is too tempting to ignore.

Undoubtedly, self-driving cars will be a strong driving force of the 4th industrial revolution. And it is more relatable to the masses than other buzzwords like AI, or blockchain.

All the major automakers are trying to bring out their own self-driving models by 2020. Last month, General Motors announced cutting 14,000 jobs to realign its focus on self-driving and electric vehicles. These changes are the new normal in the auto industry.

The seasoned automakers are facing a stiff challenge from the young and nimble tech companies such as Alphabet, Apple, Tesla, and Uber in the race to build the cars of the future.

Self-driving cars are great until everything goes downhill.

Two incidences of fatalities involving self-driving cars from Uber and Tesla in March this year brought stricter scrutiny back on the safety claims of the technology. All is not well with the self-driving technology.

On a lighter note, a scene from the very popular HBO series Silicon Valley sums up all the fears surrounding the self-driving car. In this scene a popular character, Jared gets a ride home in Peter Gregory’s driverless hybrid car.

As the funny situation unfolds the car drives itself into a shipping container, which is promptly sealed and hoisted onto a ship to a remote island. In short, Jared was carjacked by the vehicle.

Tech shopping spree

The carmakers realize that there are still serious technical and safety challenges to overcome. A quick glance at some of the biggest transactions and deals in self-driving tech will reveal that the automakers are on a shopping spree for technologies to make self-driving safer.

  • Ford $1 billion investment in Argo AI
  • Toyota $1 billion investment in Toyota Research Institute
  • GM $581 million to acquire self-driving car start-up, Cruise Automation.
  • GM $500 million investment in Lyft
  • Volvo $300 joint venture with Uber
  • Intel $250 million of additional new investments over the next two years to make fully autonomous driving a reality.
  • Uber $680 million to purchase Otto
  • Intel $15.3 billion to buy Mobileye
  • Hyundai $1.7 billion in R&D

Again, the focus shifts to Israel

From the list above, the deal that stands out is the Intel deal with Israel based Mobileye. Israel is becoming synonymous to the quintessential first stop of the tech giants for key technologies.

Headquartered in Jerusalem, Mobileye produces a vision system which could detect vehicles using only a camera and software algorithms on a processor. Due to its simplicity, Mobileye offers an effective solution to make cars better see the world while driving.

In March 2017, Intel announced their record-breaking deal to buy Mobileye for $15.3 billion. Mobileye is currently partnering BMW and Volkswagen in implementing self-driving capabilities.

The Israeli autonomous technologies do not end with Mobileye. It is only a prelude for more game-changing technologies to come. With its ideal size, availability of cutting-edge technologies, and high appetite for risk Israel is emerging as a leader in adopting autonomous vehicle technology.

A brief look at how a few of the Israeli startups are actively redefining the self-driving car story will surely blow your mind.

Phantom Auto

Let’s use the Silicon Valley reference to see how this Israeli tech company is making self-driving better. Jared had to go through all the ordeals because he did not have any control over the car to face the unforeseen circumstances.

How a self-driving car negotiates unusual traffic hazard, inclement weather, or road construction is indeed a big concern.

Luckily a Tel Aviv based company Phantom Auto, founded in 2017, has developed a teleoperation platform that enables a remote driver to take control of an autonomous vehicle if needed.

It is strictly designed as a backup measure to remotely take control of the vehicle if it faces an unforeseen scenario and gets confused or is even involved in an accident.

Phantom Auto conducted a first public demonstration of its teleoperation-as-a-service platform January 2018 on the streets of Las Vegas during CES, the massive annual tech trade show. The tech-world was amazed to see phantom steering the vehicle safely on city streets.

But a few events prompted automakers to take a closer look at Phantom. The fatal self-driving Uber accident in March 2018. This is followed in April 2018 by California updating its regulations to allow companies to test autonomous vehicles without a human safety driver if it has a remote operator who is monitoring from afar.

Driving in a crowded city can throw up challenges too overwhelming to paralyze a self-driving car. For situations like this Phantom’s teleoperation platform offers a handy complementary feature to the commuters like our poor Jared to save their day.

VayaVision

Human drivers use all their senses to gather information about the driving environment and the brain processes that information to make quick decisions to navigate the road safely.

The self-driving cars rely on sensors like camera, LiDAR, infrared, radar etc. for gathering road information. Each sensor registers an independent object and then must process the details accurately to make the correct decision.

The lack of seamless fusion of the independent sensor data causes inaccurate detections and results in a high rate of false alarms. This results in a bad riding experience.

To deal with this problem Israeli start-up VayaVision developed a raw data fusion and cognition system for self-driving vehicles. In simple terms, it works as the ‘brain’ of the car that processes the different sensor information.

The company raw data fusion technology combines camera, LiDAR, and radar input to provide a full environmental model of the driving scenario. It makes crucial decisions including lane detection, object classification and tracking, traffic and road sign recognition, and free space analysis for the car.

The VayaVision technology is making driving truly autonomous without compromising on safety and joy of riding.

Arbe Robotics

The sensor suite of autonomous vehicle works as the eyes of the vehicle. Most commonly used sensor suite comprises of cameras, LiDAR and radar. They provide redundant safety layers to the car in case one of the sensor systems fail.

But, none of these sensors is free of limitations. The traditional radar system is a functional tradeoff between medium resolution at a limited field of view and low resolution from a wide field of view. Much like our eyes, optic sensors like camera and LiDAR don’t work as well in fog or rain or snow.

In short, the currently used sensors may not operate 100 per cent of the time. This makes the vehicles vulnerable to accidents.

To solve this issue Israeli auto-tech company Arbe Robotics has developed a high-resolution radar sensor for autonomous vehicles.

Their radar system has a resolution as high as that of optical sensors, but with the level of reliability and stability of traditional radar technology

Arbe robotics’ patented system provides full mapping in four dimensions (distance, azimuth, elevation, and Doppler speed), and high resolution and long range, at an accessible price.

This technology is making the indispensable sensory suit of the self-driving car more robust and foolproof.

Hailo Technologies and Broadmann17

Self-driving cars are mobile data centers that can generate about 4 terabyte data every day as per Intel estimates. Not hard to imagine that even a few hundred such cars can easily overwhelm a network’s bandwidth.

In fact, network infrastructure in one of the biggest technical hurdles that self-driving cars must overcome before they can travel far and wide to remote places.

Inadequate network infrastructure leads to the problem of network latency- the time delay between sending information from one point to the next. Though current 4G has a latency of 50 ms, it’s not low enough to account for the split-second responses needed in self-driving cars.

The solution to the problem of bandwidth and latency lies in Edge computing. This allows efficient data processing near the source, reducing Internet bandwidth usage.

Edge computing will enable self-driving cars to respond to data almost instantaneously, as its being created, eliminating latency.

Israeli start-up Hailo Technologies developed a deep learning microprocessor to deliver data center performance to edge devices installed in autonomous cars, drones, smart home appliances etc. It can process high-resolution sensory data in real time.

Another Israeli start-up Brodmann17 develops patented deep learning technology enabling edge devices to understand their environments in real-time. It is promoted as the world’s smallest and most lightweight deep learning software-only-solution. Brodmann17 enables deep learning applications on low power processors using only a fraction of computing power, memory, and training data.

Both technologies will enable autonomous cars to make the sort of split-second decisions that human drivers make to avoid accidents and thus act as an effective autonomous driving system.

They will reduce power consumption and ensures that applications can be used effectively in remote locations. In addition, the ability to process data in the car itself adds an additional layer of security for sensitive data.

C2A Security

Modern vehicles are controlled by dozens of onboard microprocessors called Electronic Control Units controlling critical functionality including brakes, engines, steering etc. They are all connected to the data communication system of the car.

As vehicles are getting smarter, they are becoming more vulnerable to cyber-attacks. Smart vehicles in the future, will interact and learn from each other and this further enhances the risk of hackers.

Vehicles hacking is a reality, and the consequences of failure can be dangerous. But sadly, self-driving car cybersecurity has largely been an overlooked area.

C2A Security offers a comprehensive suite of cybersecurity solutions, providing in-vehicle end-to-end protection for the next generation vehicle.

Their solution provides a revolutionary safety and security layer starting from the chip level, with an effective measure to protect connected cars from malicious attacks.

The cybersecurity suite includes a patent-pending firewall for the car network, multi-network anomaly detection, microprocessor protection and diagnostics over IP infrastructure.

Imagine you are riding your self-driving car and you get a text message on the screen: “Transfer $10000 to this account or the braking system will fail”.

The scenario may seem unrealistic, but it illustrates the gravity of the cybersecurity challenges that must be overcome before autonomous and connected vehicles can be widely adopted.

And C2A Security is working towards breaking down barriers to widespread deployment of connected and automated vehicle technology.

Not there yet

Few small Israeli startups like Phantom Auto, VayaVision, Arbe Robotics, Hailo Technologies, Brodmann17, and C2A Security with their cutting-edge technologies are making the self-driving cars safer.

The technology for the autonomous vehicles is almost ready and self-driving cars are coming without a doubt. And someday in future, the steering wheel will become a part of history.

When self-driving, electrical power, and car sharing technologies will come together the commuting experience will be changed forever. This will surely have a far-reaching social impact on job security, urban landscape, habitat patterns etc.

The young generations might never need a driver’s license again. But if the society is ready for a technology as disruptive as an autonomous vehicle from the legal, logistical and civic point of view is a tough call to take.

It is only a matter of time the image of the car will change from a muscular toy to a geeky gadget. And while we wait for the fully automated vehicles to drive us around, we might want to cherish every bit of time we spend behind the steering wheel.

About the Author
Dr Rahul Dutta is a leadership fellow at the Israel Asia Center. He is currently working as a postdoctoral scientist at the Hebrew University Jerusalem. Rahul is a recipient of the prestigious DAAD fellowship offered by the German Academic Exchange Service to pursue his PhD from the Technische Universität München.
Related Topics
Related Posts
Comments