Light detection and ranging, or lidar as it is more commonly known, is seen by many to be utterly central to the future of autonomous vehicles. There are a couple of known holdouts in that regard, with Tesla adamant that a camera-based system will be sufficient, and Intel’s Mobileye wing in a similar position, but lidar startups netted a lot of investment as the automotive industry looked to solve the machine-vision problem.
Velodyne is not one of these startups, having been founded in 1983, and is something of a leader – being used in many of the self-driving car projects. Waymo is perhaps its biggest rival, having developed its own approach, but smaller rivals include Valeo, Aptiv, and Luminar. To get ahead of the pack, Velodyne has now acquired Mapper.ai, a startup specializing in mapping and localization software that is going to be used to help bundle Velodyne’s lidar offerings as a well-integrated package.
We foresee much more consolidation along these lines in the automotive industry, where the complimentary software and services are purchased and brought in-house so that they can be combined with the hardware that the OEM suppliers have spent most of their lives. This means that many startups are going to be absorbed, as will companies that thought they might be safe from the reach of an OEM – planning perhaps to be neutral but instead snared by an OEM looking to get more out of a more integrated portfolio. This is a trend that is going to be encountered across all aspects of the IoT.
Velodyne is largely seen as having invented three-dimensional lidar, in around 2005. Originally founded as an audio specialist, Velodyne Acoustics spun out Velodyne LiDAR in 2017. Its spinning lidar models, the Puck family, are frequently seen on the roofs of self-driving test cars – whirling units that spin an array of 64 laser beams out into the world and measure the returning reflections to determine what lies ahead.
However, because devices like the Puck are very expensive, the lidar industry has been racing towards cheaper options, with solid-state lidar being the most common focus area. To this end, Velodyne developed the Velarray, a unit that does not need to spin and can be fixed in place to monitor one specific orientation. Multiple Velarrays would be used in one car, with Velodyne saying that it fits perfectly behind a windshield.
As the Alpha Puck was selling for $75,000, and a new version with 128 lasers was reportedly priced at $100,000, the industry knew that it had to find a cheaper approach, and to this end, a Velarray should cost around $1,000 in volume production. The lack of moving parts is another major benefit to the solid-state design, as there have been complaints that the expensive spinning units are too fragile for these applications.
Mapper.ai is going to be put to task in Velodyne’s Valla ADAS software suite, with the entire leadership and engineering teams joining. Mapper.ai has already been using Velodyne systems in its mapping platform, capturing a view of the real world that can be understood by machines.
Mapper.ai provided a portal that allowed a customer to request a specific area to be mapped. It would then use freelance drivers to map this area, driving the route and promising to provide a result within 24-hours – hence its ‘on-demand’ branding. Once collected, the data from the cars (collected using Velodyne lidar) is converted into a semantic map, then a machine-readable format, and transformed to meet the customer’s schema or requirements. Mapper.ai can then provide updates to this map via its API.
“Velodyne has both created the market for high-fidelity automotive lidar and established itself as the leader. We have been Velodyne customers for years and have already integrated their lidar sensors into easily deployable solutions for scalable high-definition mapping,” said Dr. Nikhil Naikal, founder and CEO of Mapper, who is joining Velodyne. “We are excited to use our technology to speed up Velodyne’s lidar-centric software approach to ADAS.”
It was an interesting mix of gig-economy flexibility with bleeding-edge technological development. It is a different approach to Here and TomTom, which have their own fleets of dedicated scanning cars, although Here is using data pulled from cars on the road to provide updates in near-real-time.
The maps themselves need to be refreshed, as these ‘high definition’ maps contain a lot more than just the road layouts and street names. The maps will convey street furniture, pedestrian crossings, street signs, and more temporal obstructions that the car might need to know about, such as roadworks, diversions, or closures.
Of course, the map is not used as the sole tool to navigate, but it is a very important component of the system. The map gives the car a city-level view of the roads, and it is a view that can be enriched by V2V cooperation – with other cars sending updates to the maps. While the map might help the car plan the route from A to B, the onboard sensors are going to be responsible for the navigation itself – tracking other road users and conditioners in the immediate vicinity.
“The goal in the automotive market is to make transportation safer. By adding Vella software to our broad portfolio of lidar technology, Velodyne is poised to revolutionize ADAS performance and safety,” said Anand Gopalan, CTO at Velodyne. “Expanding our team to develop Vella is a giant step towards achieving our goal of mass producing an ADAS solution that dramatically improves roadway safety. Mapper technology gives us access to some key algorithmic elements and accelerates our development timeline. Together, our sensors and software will allow powerful lidar-based safety solutions to be available on every vehicle.”