Contributed by Bob Chabot
Autonomous Vehicle Development Hits High Gear
Fully connected and automated vehicles are expected within two decades.
As the automotive industry and tech companies push to bring self-driving vehicles to market, a handful of highly publicized accidents have highlighted the risks involved in early on-the-road testing. As the public debates the quick development and deployment of automated vehicles, a recent white paper published by Ricardo examines when these vehicles will become reality.
The driverless Sedric shared mobility concept vehicle has been making appearances at numerous international motor shows. (Image — Volkswagen Research)
Top Reason: Safety
The most frequently cited reason for development of automated vehicles is safety – vehicles driven by computers don’t get tired, distracted or impatient. Another argument is that vehicles employing optimized route planning and powertrain efficiency software are more environmentally friendly and allow drivers to use their time more efficiently.
Over the longer term, experts believe travel times could be reduced with an automated fleet by coordinating traffic flows and rerouting vehicles off overloaded streets and highways – much like an air traffic controller handling incoming and outgoing planes at an airport.
So why the hurry? Well, the shift to autonomous vehicles isn’t just about personal passenger cars. Delivery trucks, military vehicles and even taxis are primed to benefit from the connected and automated vehicle (CAV) movement.
The Ricardo white paper uses the phrase “connected and automated” instead of “autonomous,” as the authors believe it will be many years before vehicles have no driver. This description parallel’s SAE’s published hierarchy of autonomous driving.
SAE Level 0 means no assistance whatsoever: The driver must perform all tasks at all times. Level 1 is found in many vehicles with features such as cruise control and lane-keeping systems. Some vehicles today can be classified as Level 2, where the system can take over some steering and acceleration/braking tasks in certain circumstances, but most of the time the driver maintains full control.
Only a handful of Level 3 vehicles are available today. These monitor the driving environment and control steering, acceleration and braking in specific conditions. However, they return control to the driver in situations where systems do not have competence.
At Level 4, the description ‘automated’ can be applied. It allows programmed driving for the majority of likely traffic environments; however, a driver must remain on-board.
Level 5 is where the description ‘autonomous’ is fully realized. Level 5 vehicles operate remotely and without drivers. They must be capable of all tasks involved in initiating and completing a journey, whatever the conditions.
According to the paper’s authors, CAVs could be deployed on a larger scale and at autonomy Levels 4 and 5 if critical technology areas are addressed to ensure safe and effective operation. Bringing these areas to the performance level necessary will require substantial software and hardware innovations.
- Sensors are crucial for the vehicle to see its surroundings and determine how it interacts with traffic and roads. LiDAR, say the authors, is the key to real-time 3D mapping these vehicles need for spatial awareness. Currently, LiDAR systems are bulky and very expensive, but changes are expected.
- Powerful data processing of sensor output is required to identify and classify nearby objects so the vehicle can decide how to react. The CAV won’t just plan a route; the vehicle must check it every second along the way too. This will require some hefty processing hardware, as well as an adequate on-board electrical source to power it, as well as steering, braking, accelerating, etc. Oh, and the systems need to be redundant to ensure safety.
- Communication cannot be overlooked: The volume of mapping and information exchanges required is too immense to use the cloud. Fast-response functions will need to be processed on-board; regular updates can occur from the outside during a vehicle’s idle time. In addition, vehicles will be communicating with each other, pedestrians, cyclists and other road infrastructure.
- A final obstacle is cybersecurity. Just how do the automakers and OEMs protect vehicles that are interconnected, as well as connected to home and office Wi-Fi, mobile devices and anything else that needs to talk with the vehicle, such as scan tools? That one is yet to be answered.
Of all the sensor systems required for a Level 5 vehicle, the most expensive is the LiDAR for the 3D mapping of the vehicle’s surroundings. (Image — Ricardo)
Are CAVs Possible?
The push to make these vehicles is very real, and beyond the technical hurdles, it remains to be seen if society will accept the impacts CAVs will make. Automated driving, say the authors, will make vehicle ownership less appealing because of the benefits of shared travel service. This will have a lasting impact on not only the automakers and OEMs, but the repair industry as well.
The future is coming, and quickly. As industries brace for the change from the business-to-consumer to business-to-business model, repairers will need to evaluate their own place in this new ecosystem.
Want to learn more? Get the Ricardo white paper.
[Editor's note: Visit MOTOR.com for the latest diagnostic and service insights.]