Aurora

Product Navigation

  • Aurora Driver Navigation

The Aurora Driver

Created from industry-leading hardware and software, the Aurora Driver is designed to adapt to a variety of vehicle types and use cases, allowing us to deliver the benefits of self-driving across several industries, including long-haul trucking, local goods delivery, and people movement.

How it works

The Aurora Driver consists of sensors that perceive the world, software that plans a safe path through it, and the computer that powers and integrates them both with the vehicle. The Aurora Driver was designed to operate any vehicle type, from a sedan to a Class 8 truck.

Infolist Items

  • The Aurora Computer is the central hub that connects our hardware and autonomy software and enables the Aurora Driver to seamlessly integrate with every vehicle type.

    Your browser does not support the video tag.
  • Our custom-designed sensor suite—including FirstLight Lidar, long-range imaging radar, and high-resolution cameras—work together to build a 3D representation of the world, giving the Aurora Driver a 360˚ view of what’s happening around the vehicle in real time. This enables our vehicles to stop for pedestrians, travel safely on highways, avoid obstacles, and more.

    Your browser does not support the video tag.
  • Our groundbreaking software combines the best of machine learning with state-of-the-art engineering. The goal is to help the Aurora Driver to move through the world safely and predictably.

    Your browser does not support the video tag.

Aurora Computer

This central hub is the heart of the Aurora Driver—processing sensor data, running autonomy software, and allowing the Driver to make quick, precise decisions on the road.

  • computer chip icon

    Incredibly fast, reliable processing that is orders of magnitude more powerful than a typical desktop computer

  • computer icon

    Compact design allows easy integration into all vehicle platforms, from class 8 trucks to passenger vehicles

  • Thermometer icon

    Built to withstand a wide range of temperatures and on-road conditions, from Pittsburgh snow to Texas heat

A Next-Generation Sensor Suite

Infolist Items

  • Aurora believes the safest approach to self-driving perception is combining different types of sensors—high-resolution cameras, 4D imaging radar, and our custom FirstLight Lidar. By combining their unique strengths, our system is significantly more reliable than a single sensor type alone.

    Your browser does not support the video tag.
  • To travel fast, you need to see far. Our FirstLight Lidar is engineered to see more than twice the distance of even the most advanced conventional lidar systems, making it the only lidar that allows trucks to travel safely at high speeds. FirstLight eliminates virtually all interference from sunlight and other sensors—and using the Doppler effect, it can track the velocity of moving objects and people at distances and with levels of accuracy unmatched in the industry.

    Your browser does not support the video tag.
  • Currently, we also use conventional AM lidar to help the Aurora Driver identify and track objects that are closer to the vehicle. 

    Your browser does not support the video tag.
  • Cameras help the Aurora Driver recognize traffic light colors, identify stop signs, and classify objects that are far away. In combination with other sensors, they’re also particularly useful in foggy or rainy weather and at night. We’ve custom-designed the lenses, layout, and cleaning solution to ensure that they can see far enough ahead to drive safely on high-speed highways and provide a clear view all around our vehicles.

    Your browser does not support the video tag.
  • Radar helps ensure our perception system can operate in a wide range of weather conditions. We use custom imaging radar solutions that provide far greater range and resolution than traditional automotive radar.

    Your browser does not support the video tag.

Autonomy Software

Infolist Items

  • Just as humans are better drivers when they’ve been somewhere before, self-driving systems are better drivers when they learn from the collective observations of the fleet. These observations are encoded in our custom mapping and localization platform—The Aurora Atlas. These extremely detailed HD maps help the Aurora Driver plan where it’s going, and our proprietary software can determine the position of a vehicle in our fleet within 10 centimeters and 0.1 degree of accuracy.

    Your browser does not support the video tag.
  • Anyone who’s ever driven knows there are countless unique scenarios that can occur on the road at any given time. To handle inevitable edge cases we designed our perception system to “see” and track all perceived objects, even those it doesn’t recognize. We call this approach “No Measurement Left Behind,” and it’s one of the ways we ensure our vehicles always drive safely.

    Your browser does not support the video tag.
  • When developing the steering, brake, and throttle commands for our Driver, we think about creating safe, natural, and predictable movement. Our planning and control software forecasts the intention and motion of the actors around the vehicle to determine how our Driver should respond. We combine what we call engineered constraints with machine learning models to teach our Driver how to be a good citizen of the road.

    Your browser does not support the video tag.
  • As the network of vehicles powered by the Aurora Driver grows, the road networks, traffic scenarios, and weather conditions it can handle will expand. To deploy our Driver safely and quickly in new operational domains, we’re investing in teleassist, a set of technologies that allows remote specialists to review contextual information like video feeds and point the Aurora Driver in the right direction. This helps us avoid hiccups and get our vehicles back on track quickly, meaning more rides, faster deliveries, and more satisfied customers.

Our virtual testing suite

Infolist Items

  • Simulation allows us to rapidly and safely develop the capabilities and robustness of the Aurora Driver. We can be confident the Aurora Driver will perform new maneuvers safely on the road when it has effectively already practiced them millions of times in simulation. In just one hour, our Virtual Testing Suite can simulate the equivalent of over 50,000 trucks operating on the road.

    Your browser does not support the video tag.
  • We don’t leave testing to the end of the development process. As code is written, we write both unit tests (e.g., seeing if a method to calculate velocity gives the right answer) and integration tests (e.g., seeing whether that same method works well with other parts of the system). New work must pass all relevant tests before we can merge it with the larger codebase, allowing us to identify and fix issues early.

  • We use real-world log data to create a series of specialized perception tests. For example, say a bicyclist passes very close to one of our vehicles. Specialists would review that footage and then label things like object category (bicyclist), velocity (3 mph), etc. We would then use that “ground truth” to test how well new versions of perception software can determine what really happened on the road.

    Some tests assess how well the system performs across the board. For example, “How many pedestrians does the system correctly identify?” The rest evaluate specific capabilities, like tracking close objects. For example, “Does the perception system see this bicyclist right next to our vehicle?”

    Your browser does not support the video tag.

On-Road Testing

We’ve adopted a “smarter, not farther” approach to on-road testing. Instead of blindly pushing to drive more and more miles, we’ve continued to focus on collecting quality real-world data and on getting the most value out of every data point. For example, we amplify the impact of real-world experience by flagging interesting or novel events and incorporating them into our Virtual Testing Suite. One real-life situation can inspire tens or even hundreds of test permutations, all of which can be continually used to fine-tune existing capabilities. The Online to Offline pipeline allows us to rapidly convert on-road events into virtual tests.

The Aurora Driver in action

Watch the Aurora Driver slow down and wait while a pedestrian gets into a large truck before smoothly nudging around.

Watch how during this unprotected turn, the pedestrian acts like they are going to step out into the crosswalk. Immediately, the Aurora Driver plans to yield (you can see how the green carpet shortens). Then, as soon as the pedestrian changes direction away from the crosswalk, the Aurora Driver smoothly resumes its trajectory and completes the turn.

Lots of oncoming traffic, lots of pedestrians in the crosswalk, and a smooth unprotected left turn.

Watch the Aurora Driver slow down and complete an unprotected right turn while yielding to pedestrians.

The Aurora Driver completes a very comfortable lane change on a highway, moving at 50mph.

Aurora

Footer Navigation

  • About

  • Company Company
  • Careers Careers
  • Investors Investors
  • Press Press
  • Blog Blog
  • Explore

  • Our service Our service
  • Aurora Driver Aurora Driver
  • Safety Safety

Sign Up

Email confirmed!

Follow Us

Social Navigation

All Rights Reserved © 2021 Aurora Innovation Inc.

Privacy Policy Terms of Service Security