Autonomous driving systems maker Amigray, Inc. promises to turn ordinary passenger cars into driverless vehicles by installing artificial intelligence (AI)-powered systems for a fraction of the cost of a robo-taxi.

“When you’ve built a software and hardware stack that costs $100,000 per vehicle, it’s not a business case to fit it into a passenger vehicle that costs around $30,000. This expensive system makes sense for robo-taxis. can run 12-14 hours per day because you eliminate the need for drivers, who make up a significant portion of the overall cost,” said Erin Ophir, CEO of Amigre, from its Israel-based office. told Minute in an interview. “In a passenger vehicle, though, you need something that costs an extra $2,000 to $3,000.”

Instead of relying on high-definition (HD) maps, also known as 3D maps, which many automakers commonly use in driverless cars, vehicles powered by Amigray’s AI systems can use real-time visual information. Interprets — such as objects, lanes, traffic lights, and pedestrians — to make driving decisions. Most driverless cars use HD maps in conjunction with Light Detection and Ranging (LiDAR) systems and radar.

While maps provide a static, detailed road environment, LiDARs and radars offer real-time, dynamic data on objects and obstacles. Sensor data provides real-time updates. Waymo’s autonomous vehicles, for example, use a combination of HD maps, LiDAR, and radar to navigate complex urban settings with high precision. However, Elon Musk-led Tesla and UK-based startup Wave do not use external HD maps. Instead, they use camera-based vision, deep neural networks, and real-time sensor data to enable cars to learn in real-time like humans.

HD map-based systems offer high accuracy, context, and better predictability, but they are expensive, require constant updates, and are prone to cyber-attacks. Mapleless systems rely on real-time data from sensors, making them cost-effective, adaptable to real-time road changes, and scalable. But they can struggle with complex environments and lack predictive capabilities without prior map data.

Despite the debate over the trade-offs in each approach, autonomous driving system maker Amegri, Inc. is betting on the latter approach. According to Ophir, its autonomous driving systems for cars and even buses are primarily powered by AI.

This “bio-inspired design” relies on two basic components: perception and movement planning. Ophir explained that “perception mimics human vision by interpreting objects and environments, while movement planning mimics cognitive processes in the brain’s cortex.”

In India, object detection systems must be trained with specific images, such as cows, to ensure that they understand the need to produce them on roads.

The system uses video feeds from multiple cameras, and processes this data through a series of “specialized deep convolutional neural networks (CNNs)”. CNN is a deep learning (a subset of machine learning, which is an AI technology) model that analyzes visual data by recognizing patterns and features in images. In driverless vehicles, CNNs process camera feeds to detect objects such as pedestrians, traffic signs, and other vehicles.

According to Ophir, each network is trained on millions of images to detect and classify various objects, such as traffic lights, signs, lanes, pedestrians, and vehicles, within a 300-meter radius. Covers a 360-degree view.

“This allows the system to create a real-time, HD-equivalent (3D) map of the environment, which is then integrated into a motion planning component that learns over time and a process called learning by imitation. Adopts through. The child learns,” Ophir explained. The system can be integrated with LiDARs and radars if needed, he added.

Also read: Meet the four musketeers of India’s quantum computing dream.

According to Ophir, many robotic taxi services that use HD maps “get confused if road conditions deviate from their programmed expectations.” HD maps also require constant high-bandwidth communication, which can present challenges in network congestion, rural locations, or tunneled areas. Where signals can be weak or nonexistent, HD maps are expensive to maintain and can expose vehicles to cybersecurity risks, according to Ophir.

See full image

Fortune Business Insights predicts that the global autonomous vehicle market size, which was valued at $1.5 trillion in 2022, is expected to reach $13.63 trillion by 2030. According to Ophir, Amigre has already signed a 10-year contract with Continental to develop a software. According to Ophir, a vehicle platform that incorporates their neural network-based motion planning. This allows vehicles to gradually gain autonomous driving capabilities through over-the-air software updates. Starting with features like autonomous parking and traffic jam assistance, the system aims to eventually achieve full self-driving functionality.

“Furthermore, our system is hardware agnostic, allowing customers—including Tier 1 suppliers and OEMs (Original Equipment Manufacturers)—to choose their preferred hardware without being tied to specific components,” Ofir said. . He added that this hardware independence is particularly beneficial because it addresses the diverse needs of different vehicle price segments—from entry-level to high-end models, which often have more computing capabilities and sensors. The configuration is completely different.

Autonomous bus

From a safety perspective, Amigray is currently focusing on passenger cars with “level 3 autonomy,” where the vehicle can drive itself, but the driver remains in charge. The company is also focusing on L4 driverless electric buses. These vehicles do not require human interaction in most situations, but a human operator can. Manually override the system for security reasons.which can cause the AI ​​model to make mistakes or delay decisions.

Autonomous buses, according to Ophir, are “easier to implement than passenger vehicles because they operate in geo-fenced areas and follow fixed routes, allowing for immediate information about the environment.” Despite the potential for faster deployment, strict regulations govern the introduction of autonomous buses, including passing stringent security tests and cyber assessments before allowing passengers to ride, he added.

According to Ophir, there is a worldwide shortage of bus drivers—currently at 15% and growing—which puts pressure on public transport operators. Because of this shortfall, there are pilots for autonomous buses in 22 markets around the world, including Japan, where the government plans to deploy 50 autonomous buses by 2025 and 100 by 2027.

The economic implications for operators are significant, Ophir says, estimating that autonomous buses could save between $40,000 and $70,000 per vehicle annually. “This can drastically improve operational margins and service levels. For example, a partnership with a leading public transport operator with a fleet of 70,000 buses is an example. The transformative potential of this technology” said Afir.

Adaptation and customization

Amigray’s AI systems can also operate in diverse environments, including the congested urban areas of Israel and Tokyo, “adapting to driving on both sides of the road, a feat not typically achieved by other autonomous driving companies.” “. He explained that for a vehicle to operate in a new area, the entire system needs time to learn the local traffic patterns and road characteristics.

While most traffic signs are the same across countries, there are unique signs or rules that need to be programmed into the system. For example, when transferring from the US to Germany, the software must be updated to reflect local laws, such as the prohibition of right turns on red.

Although there are no driverless cars in India, Ophir believes that the main challenge of driving such cars “is not congestion, because the system can adapt to traffic jams, as found in Indian cities like Bengaluru or Pune.” “. Instead, the focus is on training the software to recognize and respond to local conditions. “For example, in India, object detection systems must be trained with specific images, such as cows, to ensure that they understand the need to produce them on roads. Ophir concluded.