There is little doubt that the DARPA Grand Challenge and DARPA Urban Challenge autonomous vehicle competitions have driven unmanned systems innovation. In particular, the U.S. government-sponsored challenges have motivated companies to develop innovative core technologies that will be used in future generations of autonomous vehicles. Velodyne Lidar Inc.’s HDL-64E LIDAR (Light Detection and Ranging) scanner and GrayMatter Inc.’s Visionary point cloud visualization software are two examples of such core technologies. The two companies have now formed a partnership that should drive commercial acceptance of the Velodyne LIDAR solution, as well as foster the development of new autonomous vehicle solutions.
Velodyne’s Unique Approach
Morgan Hill, Calif.-based Velodyne has historically been known for its speaker systems. In 2004, Velodyne founders entered the DARPA Grand Challenge-an autonomous vehicle race through the Nevada desert sponsored by the U.S. government’s Defense Advanced Research Projects Agency (DARPA)-as Team DAD (Digital Audio Drive). The vehicle traveled only 6.2 miles before it stalled. After failing to complete the second Grand Challenge in 2005, Velodyne dropped out of the DARPA competitions to focus on developing a commercial version of a unique LIDAR technology that constructs a 3D representation of the environment surrounding sensor-equipped vehicles. Within a year, Velodyne introduced the HDL-64E LIDAR system.
Prior to Velodyne’s entry to the market, the most popular laser scanning systems were single-laser diode sensors, coming from companies such as SICK and Ibeo, which offered a 180-degree field of view. Multiple sensors were required for a 360-degree view, or units had to be actuated for a vertical sweep.
Velodyne overcame the requirement for multiple laser scanning sensors by developing a spinning unit containing a 64-element array of laser diodes, which allows a single sensor to sweep a full 360-degree horizontal, 26-degree vertical, toroid. The Velodyne sensor boasts a higher resolution compared with competitive products, producing 1 million data points in a second versus the thousands generated by other sensors.
For the DARPA Urban Challenge, held in 2007, many teams made use of the new Velodyne LIDAR sensor. The product was a success-four of the top five teams completing the urban course used the HDL-64E. It was easy to determine which teams used the product. One team, for example, used only two LIDAR units on its vehicle, while another team, still using the popular SICK sensor, had 10 units on its entry.
It was during the 2007 challenge that Velodyne began working with Metairie, La.-based GrayMatter. GrayMatter was another former DARPA Challenge team that created a robust autonomous vehicle system called GrayMatter AVS. AVS is a hardware/software control system for autonomous vehicles that can be linked into a number of commercial sensor and GPS systems. The company also offers its Visionary point cloud visualization software. With a solid understanding of system integration on its own vehicle, GrayMatter was poised to be a valuable partner for Velodyne.
The result of the initial collaboration was to have the GrayMatter Visionary point cloud visualization software interpret data from the Velodyne LIDAR and plot it to create a visual representation of the environment. With this level of visualization it is possible to detect environmental features in real time, leading to capabilities like road edge detection, obstacle detection, and safety features. In addition to enabling cutting-edge autonomous ground vehicles, this type of information is also used in adaptive cruise control systems on the more advanced cars on the commercial market.
A new partnership between GrayMatter and Velodyne announced last May not only formalized the HDL-64E/Visionary integration, but also created a seamless distribution model, so that Velodyne sensors now ship with a basic trial version of GrayMatter’s software. It also allowed GrayMatter to build additional functionality into its software package. Now when customers purchase Velodyne’s HDL-64E sensor, they can implement complex behaviors like obstacle avoidance right out of the box.
System integration is typically the most difficult, time-consuming component of robot development. Although numerous commercial sensors and subsystems for robotic application are available for use, difficulties arise during component integration and when behavioral algorithms that act upon sensor-generated data are being developed.
For researchers and academics, the complex integration might only necessitate extra hours on the part of graduate students. For commercial entities, however, lengthy integration work implies increased costs. An advanced sensor solution that ships with most integration and behavior software already written frees developers to concentrate on other autonomous behaviors, thereby reducing overall systems development costs, as well as increasing the pace of innovation in autonomous vehicle development. This is exactly what the Velodyne/GrayMatter partnership provides.
Enhancing Velodyne’s Value Proposition
By integrating the GrayMatter Visionary software with its LIDAR hardware, Velodyne increases the value proposition of its overall sensor solution. Good thing, as the Velodyne sensor is currently priced around $75,000, a huge increase over less advanced LIDAR units that cost approximately $5,000 to $10,000. Still, the Velodyne solution is a functional replacement for multiple LIDAR units from its competitors, and therefore can be cost justified to some degree.
The argument for reduced number of LIDAR systems aside, the business case for Velodyne’s solution can be strengthened if its prices are reduced. To do so, the company must prove the utility of its product in commercial environments. If it is successful in these efforts, then Velodyne can produce its LIDAR units in higher volumes and drive down the product’s price. The incorporation of the GrayMatter software makes the Velodyne LIDAR solution more compelling and should result in the company gaining a stronger hold in the market.Read More