How Meridian puts the drive in the self-driving Olli
Since Local Motors debuted its first self-driving vehicle, Olli, a month ago, we’ve told you about the team taking the vehicle out into the market, but there are dozens of other people who have been integral to the product’s early success. Not all of those people work directly for Local Motors.
Many times during the development of a vehicle, suppliers and partners go unrecognized, but in reality, a strategic partner can carry much of the burden as well. While Local Motors has a great internal team, we also want to give kudos to our external partners and their team members.
Meridian Autonomous Inc. and its team have been the brains behind mapping, interpreting sensor data, and routing that information to the various components that steer, accelerate and stop Olli. Simply said, Meridian has provided much of the drive behind this self-driving vehicle.
Olli operates by interpreting the physical world through a set of redundant sensors. The vehicle and all of its high-tech gadgets, electronics, and steering system are nothing without the software that ties everything together, allowing the vehicle to safely maneuver city streets.
Before Meridian gets to work on programming the vehicle, they first map the streets on which the vehicle will operate by using lidar technology. Meridian drives the routes with a lidar (light detection and ranging) sensor attached to the roof of a normal, manually driven vehicle. Lidar is a very accurate tool used to identify and localize objects. It shoots out millions of data points in all directions that is uses to create a 3D map of the environment, and is accurate to the half inch.
During this process “landmarks” are designated. These are fixed points (i.e. buildings, light posts, etc.) that the self-driving vehicle can reference during operation. The map created from this process will be embedded onboard the vehicle as well as available to a fleet manager to define routes and stops.
Now for the part that makes the vehicle truly autonomous: The system architecture is defined by separate and specific layers. The high-level layer interprets the sensor data; everything from lidar, to optical cameras, to bumper switches. This layer is for localization, obstacle detection and route optimization. The low-level layer is all about control of the vehicle. It is responsible for operational functions such as closing the door, accelerating, and braking.
A wholly separate layer deals with emergency situations and is not accessible by any outside connection. This way, no unauthorized parties can interrupt its chain of command. If the high-level layer detects an issue the safety chain is activated and the vehicle will come to a safe stop before commencing a troubleshooting sequence involving a human, remote supervisor.
Between redundant sensors, safety layers, and low-speed operation, fully autonomous vehicles are a reality today. We are excited to continue to pioneer this field with our incredible partners. Stay tuned to hear more about Meridian Autonomous Inc. and the exciting projects they have on deck.
And if you want to experience autonomous drive for yourself, we plan to have Olli up and running in National Harbor, Maryland in September. Additionally, Olli will be piloted in both Las Vegas and Miami-Dade County in the fourth quarter of this year.