Autonomous Driving: The Future Of Trucking
By Owen Liu
The dream of autonomous driving has gotten closer thanks to improvements in artificial intelligence, navigation and sensor technology.
However, it will still be some time before we can sit in a fully autonomous vehicle. Just as we did not achieve the moon landing through a Day 1 launch of Apollo 11, it will take a similar process of trial and error to produce fully autonomous vehicles.
Over the years we moved forward through many small steps to stand on the cusp of that giant leap of driving technology. And, because autonomous driving was the product of many different actors, there are a variety of solutions involved, all independently seeking to push the autonomous driving to the next level.
Autonomous driving currently exists at a number of ‘levels’ based on the complexity of the autonomous systems.
The first level of autonomous driving vehicles are equipped with adaptive cruise control and lane keep assistance. These tools do not free drivers from actually driving, but help relieve driving fatigue.
The next level of automation further assists by controlling speed and steering. It maintains the distance from the vehicle in the front, and steers to the center of the lane. As with the first level, the driver must remain engaged and monitor the environment at all times.
The third level of automation permits autonomous driving in ideal conditions, such as divided highways at certain speeds. The vehicle can be driven hands free, although drivers are still required to use the wheel when driving conditions are less than ideal.
At the fourth level, the vehicle is capable of performing all driving functions under known conditions.
Level five automation, which does not currently exist, would be a true driverless car that can maneuver though all road conditions without human intervention. There could be no steering wheel or pedals and you could play poker with your friends, with your undivided attention on the table.
The foreseeable demand for autonomous driving is challenging car manufacturers to overhaul their thinking about the entire driving system. The auto industry used to focus on components and specific devices or subsystems. They relied on suppliers to provide a particular device that fit the vehicle, so the supplier had to do the system thinking for the car manufacturer.
However, in the wake of demand for future autonomous driving, auto makers are having to lead the design of a platform of integrated sensors, computational hardware, networking, and software. Different real time systems must operate together, including environment perception, localization, planning and control.
New challenge: a comprehensive view
This new challenge requires a comprehensive view of the entire system. While the level five driverless vehicle is still being pursued, at the same time people are coming up with solutions that could be practically feasible today.
The trucking industry is paying particularly close attention to these solutions. The truck driving profession is known as a challenging occupation due to both its physical demands and lifestyle sacrifices. Given the shortage of truck drivers in the current marketplace, autonomous driving truck, when available, would be a perfect solution. Any solution less than perfect, but practically achievable in the near future, is still highly desired. Here are two examples on the horizon.
One venture company “A” combines an onboard self-driving system with a human remote operator standing by to guide the truck when necessary. It is not a completely autonomous solution. Instead, the remote operator watches over the autonomous driving in a remote control room.
Cameras mounted on the truck provide real-time visual information to the control monitors. When the truck is driving on a highly controlled environment, such as interstate highways, the automated driving systems take over. Typically, a truck spends most of its time on highway in a long-haul assignment.
When the truck enters into a less controlled environment, such as local roads to pick up loads at a warehouse, the remote operator takes over and steers the truck through a video-game-like interface.
Making a truck drive without a person in it creates a number of benefits. Multiple trucks can be operated by a single remote operator. Standardized shifts, rest room breaks and physical stretching are all much easier now. Most importantly, it is extremely appealing for drivers to be able to go home to their families at the end of the day.
The other venture company “B” takes a different approach by doubling productivity, having one driver steer two trucks at once. While much technology research focuses on autonomous driving, B is focusing on autonomous following. The driver drives in the lead truck. Every action he takes is recorded, such as steer maneuver, throttle application or brake application. The recorded information is then transmitted to the computer in the following truck.
The following truck does not have a human driver. It uses the information from the lead truck to tail closely behind. This approach significantly reduces the technical challenge required for full autonomous driving. The productivity can be tripled, quadrupled, or more if more trucks can follow the lead truck.
New technology tends to rewrite industry rules. The realization of autonomous driving in the near future will significantly reduce the cost of operation for trucking companies. As expert advisors for business owners, Mazars navigates the next step of the industry for our clients so they can stay on top of the competition, and maximize profits and returns. We look forward to discussing the exciting changes in technology for trucking and how it can create opportunity for your business.