Nexar Broadens Mapping Capabilities for AVs, Robotaxis and ADAS with New Mapping Layer

 Nexar, a leading AI computer vision and mapping company has announced a new mapping layer that provides an understanding of expected driving etiquette vs. actual driver behavior. Based on aggregated data acquired through the company’s extensive network of 700,000 dash cams, Nexar highlights the gap between everyday driving scenarios compared to the official rules. By providing automated driving systems with this context, driver behavioral maps can contribute to improved safety and better interaction with human drivers.

Recommended: Enhancing AI: Why New Technology Must Include Diversity

The advancement of automated driving technology faces numerous challenges in real-world environments. For example, the interaction of automated driving technology with human drivers, pedestrians and cyclists can prove complex, as they may not always follow official rules. To offer a solution, Nexar’s new mapping layer has been introduced to complement existing layers. All layers of maps are crucial for a complete driving experience. From the traditional “Base Map” to the more detailed “Lane Semantics” and “Road Furniture” layers, each one adds a unique level of detail.

Nexar is taking things one step further by identifying scenarios where human driver behavior diverges from the official rules of the road, impacting autonomous vehicles that lack intuition and contextual knowledge. Key examples include virtual stop-lines and virtual crosswalks, which occur when obstacles or common jaywalking lead to an increase in decelerations and full stops. Other scenarios involve turning lanes, where drivers overwhelmingly choose to turn at an intersection, and speed limits where drivers typically drive well above or below the posted limit. Lastly, school zones have behavior variability around drop-off and pick-up times during the week. These differences can greatly impact how autonomous vehicles navigate these areas.

Nexar’s comprehensive findings highlight the importance of real-time driver behavioral maps in enhancing both AV and ADAS performance, making roads safer and more efficient. By using real-time data, driver behavioral maps can provide a thorough understanding of patterns that impact autonomous vehicles. This includes being able to identify anomalous driving behaviors, such as hard braking and aggressive lane-cutting, and map them in real-time. For example, Nexar has uncovered that there is more than double the chance for a driver to perform a hard brake on a highway exit than on the highway itself, and drivers are 14 times more likely to perform hard brakes on residential roads than on highways. Another example is that in specific sections of Florida’s Turnpike in Miami, nearly 90% of drivers exceeded the 70 mph speed limit by over 10 mph. This is in contrast with highways in Miami overall, where drivers went more than 10 mph above the speed limit only 6% of the time.

Recommended: Enhancing AI: Why New Technology Must Include Diversity

“Our work explores the essential role of driver behavioral maps in deepening our comprehension of human driving habits and the unwritten rules governing the road,” said Ilan Kadar, VP of AI & Automotive at Nexar. “This new mapping layer, generated from Nexar’s network, acts as a significant enhancement to our mapping solutions, facilitating safer advanced driving assistance systems as they integrate into our everyday commutes.”

Latest Insights: Synthetic Data: A Game-Changer for Marketers or Just Another Fad?

[To share your insights with us, please write to sghosh@martechseries.com]