SAN JOSE, Calif. – Ouster, a San Francisco-based developer of high performance LIDAR technology, announced today at the NVIDIA GPU Technology Conference (GTC) in San Jose two new products and the adoption of the NVIDIA DRIVEâ„¢ AI platform for its R&D efforts. The new products, the OS-2-64 and OS-1-16, extend the company’s performance and pricing leadership in LIDAR, with sensors that lead in virtually every LIDAR use-case. Ouster’s use of NVIDIA DRIVE will streamline AV and ADAS development and integration of LIDAR sensors.
OS-2: Extended range
Ouster is driving performance further with the OS-2 LIDAR sensor. The OS-2-64 delivers 200 meter range with 64 beams at a tight spacing of 15.8 degrees, providing tailored performance for long-range sensing for highway AV and ADAS applications. The OS-2 is roughly 50% larger than the OS-1 form factor “ comparable in size to other small form factor LIDAR sensors on the market at roughly 110mm diameter x 85mm height, and a mass of around 600 grams.
Compared to other extended range sensors on the market, the OS-2 doubles the beam count and extends range further, all at a significantly lower price of $24,000 per unit.
Samples of the OS-2-64 will be available for customers in early Q3 2018, with commercial production later in Q3.
OS-1-16: Budget-friendly
Ouster is expanding the OS-1 family, introducing the OS-1-16, a 16-channel LIDAR. The OS-1-16 has the same specifications as the OS-1-64, but with a lower beam count “ for the industry-leading price of $3,500 per unit. Ouster offers unparalleled flexibility in OS-1-16 configurations, offering three versions of the OS-1-16, including an option to customize:
- OS-1-16A “ uniform beam spacing over the 31.6° vertical field of view
- OS-1-16B “ higher beam density around the horizon
- OS-1-16C “ higher beam density below the horizon
The OS-1-16 is available immediately in limited quantities, with broad availability in May 2018.
With a range resolution of 1.5 cm, the OS-1-16 is the ideal tool for drone navigation and surveying applications. The OS-1-16, benefits from the same improvements over incumbent technology as the OS-1-64, and OS-2-64, including low beam divergence (>2.5x smaller spot sizes/better spatial resolution) and angular accuracy down to the thousandths of a degree.
“With the launch of the OS-2-64 and the OS-1-16, alongside our first product, the OS-1-64, Ouster solidifies its position as having the best LIDAR sensors for every application,” said Ouster CEO and co-founder Angus Pacala. “Ouster is committed to being the performance and price leader without compromise. We’re listening to customers and working hard to supply what the market needs. We’re excited to see what new applications our sensors enable.”
NVIDIA DRIVE
Joining the NVIDIA automotive ecosystem, Ouster will incorporate the NVIDIA DRIVE platform to accelerate processing of LiDAR data. This AI supercomputer combines deep learning, sensor fusion and surround sensor data to accurately create a full, 360-degree environment and produce a robust representation of static and dynamic objects surrounding the vehicle.
“The data generated by LIDAR requires an enormous amount of computational horsepower to understand the environment surrounding a vehicle,” said Glenn Schuster, Senior Director, Technical Marketing, NVIDIA. “With NVIDIA DRIVE, Ouster can leverage our AI supercomputer and the vast ecosystem of partners developing autonomous vehicles”
Launch of partner program
Finally, Ouster has formally launched its partner program. Interested distributors and integrators can email [email protected] to be a partner and join the company’s global network.
About Ouster
Ouster is a leading developer of LIDAR and perception technology for the autonomous vehicle and robotics sectors. The company’s flagship product families, the OS-1 and OS-2, provide industry-leading performance, scalability, reliability, availability, and form factor. Ouster’s corporate headquarters and manufacturing facility are located in the Mission District of San Francisco. For more information, visit www.ouster.io.
