Chris Brewer, chief program engineer for Ford Autonomous Vehicle Development, said in a blog post that the new vehicles employ the previous platform, "but up the processing power with new computer hardware" and that the "electrical controls are closer to production-ready".
Having expanded the number of self-driving Fusions in its fleet to around 30 this year, the company plans on tripling that again by the end of 2017. These vehicles have seen tremendous technological advancements, which include the ability of driving on roads covered with snow and also being fully functional in the dark without headlights. Brewer says the cars' two new lidar sensors, down from four, gather as much data but are sleeker and more efficient.
The problem though seems to be the computer power to process the data from those sensors. In August of this year, Ford revealed its intention of releasing a high-volume, completely autonomous fleet in 2021.
If automakers want to sell self-driving vehicles, they need to work on sensor integration and can't have vehicles looking like they bolted a bunch of cameras and LiDAR sensors on top of a ski rack, which is how most test vehicles look like these days.
Ravens' Steve Smith Likely Playing His Final Game
Prior to last season, Smith said he would retire once the season ended, regardless of where the Ravens finished. Smith is still respected by many in the organization, not the least of whom is owner Jerry Richardson.
The car's updates are focused on its virtual driver system, with the vehicle's brain located in the trunk.
The new vehicle will test out two new hockey-puck-sized LiDAR sensors, which generate millions of beams from the car's front to produce a 360-degree view of the auto. The company will debut the vehicle at next month's CES conference in Las Vegas, and it'll be on display the following week at the Detroit auto show. "These new sensors possess a sensing range roughly the length of two football fields in every direction", added Brewer. Short- and long-range radar sensors help the auto "see" objects through heavy rain, fog and snow and also detect moving objects. This built-in supercomputer can generate a full terabyte of data per hour, Brewer says.
The objective is to have an automobile capable of Level 4 autonomy - that is, capable of handling typical driving conditions, if not out-of-the-ordinary conditions - available for commercial ride-hailing or ride-hailing fleet services by 2021.