Embracing the Future: 5 Advantages of Innoviz’s Perception Software

  • October 15, 2024
  • Assaf Shabat, Product Manager
perception image for a blog post

A key element of autonomous vehicle technology is perception software, which is responsible for interpreting and understanding the vehicle’s surroundings. Innoviz’s Perception Software utilizes advanced sensors, such as LiDAR, to create a comprehensive 3D map of the environment, identifying and classifying various objects, obstacles, and road features. This real-time understanding of the vehicle’s context is essential for enabling safe and reliable autonomous driving capabilities.

Innoviz Technologies is at the forefront of the autonomous driving revolution, and its pioneering Perception Software is redefining the capabilities of advanced driver assistance systems (ADAS). This cutting-edge solution, seamlessly integrated with Innoviz’s industry-leading LiDAR sensors, delivers five key advantages that are transforming the way vehicles perceive and navigate their surroundings. Here they are:

1. Seamless Integration with Innoviz’s LiDARs

Innoviz’s Perception Software is one of the only solutions of its kind that enables Level 3 autonomy and is currently installed on a mass-produced consumer vehicle – the BMW i7, available in Germany. This integration with Innoviz’s LiDAR sets a new standard for ADAS capabilities in the automotive, being one of the first to achieve high availability in the context of mass automotive production while simultaneously ensuring smooth collaboration between OM and CV teams.

Another significant advantage of choosing Innoviz is the ability for customers to acquire both the LiDAR hardware and the complementary Perception Software from a single vendor. Clients can benefit from streamlined integration services, and simplified technical support, as they have a single point of contact for all their needs, from hardware inquiries to software updates.

2. Enhanced Object Detection and Classification

Innoviz’s sophisticated computer vision and deep learning algorithms enable its Perception Software to accurately identify and categorize a wide range of elements, including vehicles, pedestrians, cyclists, and static obstacles. This advanced perception capability is crucial for being able to provide safe and reliable autonomous driving at Level 3, even at high speeds of up to 130 kmph.

lidar perception software  image for a blog post

In addition to its advanced object recognition capabilities, Innoviz’s Perception Software excels at accurately detecting and interpreting lane markings on the road. This critical functionality ensures that autonomous vehicles can seamlessly navigate and maintain their position within their lane, further enhancing safety and driving performance.

3. PC Classification Plus

Innoviz’s Perception Software stands out with its unique PC+ feature, which goes beyond to provide sophisticated collision potential assessment on the pixel level. This advanced capability allows the system to differentiate between objects that pose a genuine collision risk and those that don’t.

For instance, PC+ can distinguish between a small obstacle, which represents a potential hazard for collision, and a splash of water from a moving vehicle, which doesn’t pose a direct threat. This nuanced understanding of the environment significantly reduces false positives and unnecessary interventions, leading to smoother, more reliable autonomous driving experiences. Thanks to the same mechanism, in harsh weather such as snowfall, the PC+ feature classifies snowflakes as ‘drivable,’ allowing the vehicle to navigate through them safely.

We’ve designed our Perception Software to also excel in various road and vehicle conditions (passenger location, gas status, etc.). This adaptability enables consistent performance even in challenging conditions.

4. Advanced Blockage and Range Estimation

Innoviz’s Perception Software can differentiate between temporary obstructions that can be easily addressed, such as dirt that requires cleaning, and water droplets that mostly disappear on their own. This granular level of diagnosis enables drivers or autonomous systems to take appropriate action promptly.

lidar perception software image for a blog post

In the case that the range estimation feature indicates degradation, the system will also include a range degradation estimation feature. This unique capability evaluates precisely how environmental factors affect the sensor’s range estimation performance. For instance, in adverse weather conditions such as heavy rain, the system can quantify the extent to which this impacts LiDAR’s ability to accurately measure distances. By continuously assessing and reporting on range degradation, the software enables autonomous driving systems to adapt their behavior in real time.

5. Extensive LiDAR Validation and Testing

At Innoviz, we are committed to ongoing research and development, ensuring that our software remains at the cutting edge of AI and perception technology. We regularly deploy updates to enhance performance, add new features, and address any emerging challenges in the autonomous driving landscape.

In April 2024, our Munich team embarked on an 18-day winter test drive across Western Europe, putting the InnovizTwo™ LiDAR and Perception Software to the test in harsh winter weather conditions. This extensive field testing not only validated our technology’s performance but also gathered valuable data for ongoing learning and development, ensuring our solutions remain at the forefront of autonomous driving technology.

Driving the Future of LiDAR Autonomous Perception

Through these key advantages, Innoviz’s Perception Software represents a high-performance ADAS solution that is poised to redefine the driving experience. As the automotive industry continues to evolve, Innoviz remains at the forefront, shaping tomorrow’s smart and autonomous vehicles with its cutting-edge technology.

To learn more about Innoviz’s Perception Software, please get in touch with our Sales team by filling out this form.