Billions of dollars have been invested in autonomous vehicle (AV) technology, and the industry has spectacular advancements to show for it. However, we are still nowhere near achieving widespread deployment of fully autonomous vehicles.
At this time, no companies offer a Level 5, fully autonomous vehicle that can operate in all circumstances with no human intervention.
AV leaders have achieved Level 4 technology— most driving functions are automated, but a human is still needed in unusual circumstances. These vehicles, mainly campus shuttles or employee buses, operate only along pre-defined routes under specific circumstances (daytime, good weather).
Self-driving cars available for the public to buy for personal use are still firmly at Level 3. In a Level 3 vehicle, a computer handles two or more simultaneous driving functions, such as cruise control and lane keeping. Cadillac’s Super Cruise system is a widely recognized leader at this level. Consumer Reports ranked it first in a 2018 comparison test between Cadillac’s Super Cruise, Nissan/Infiniti’s ProPilot Assist, Tesla’s Autopilot, and Volvo’s Pilot Assist systems.
In order to make it possible for a self-driving car to navigate, an autonomous driving system is built on a collection of software and sensors. These elements work together to create a living map of the world.
While most AV companies rely on the same basic technological foundations of lidar, radar, cameras, and ultrasonic sensors, there is debate on whether lidar is necessary in cases where radar and cameras are already used. Many companies use lidar along with cameras and ultrasonic sensors to form multi-layered safety nets.
However, Tesla and Nissan are vocal opponents of lidar. Both companies rely solely on cameras and ultrasonic sensors for their autonomous systems.
According to Nissan General Manager Tetsuya Iijima, “At the moment, lidar lacks the capabilities to exceed the capabilities of the latest technology in radar and camera.”
Machine learning and artificial intelligence (AI) enable companies to construct simulation environments to quickly test rare scenarios and hardware tweaks on their cars. This virtual world is valuable because it has incredibly detailed mapping of intersections, routes, and even whole cities.
Currently, a small handful of startups is exploring Vehicle to Everything (V2X) communication. These communication systems go beyond all vehicles speaking to each other. VX2 systems would enable traffic lights to talk to cars, which in turn talk to other cars, which in turn talk to weather services, and so on. These incredibly dynamic systems could prevent many accidents and traffic jams.
However, VX2’s barriers to entry are high. Adopting the technology would make every infrastructure project, traffic light, and new car much more expensive. Until these sensors and communication devices experience a significant drop in price, V2X technology is unlikely to be adopted outside of special zones or forward-thinking cities willing to make significant investments.
We are more likely to experience a combination of Level 4 and 5 automation within strict confines, such as shuttle and delivery services, than we are to see Level 5 automation in every condition and on every road. Good news— it’s not a total loss. Even limited automation would relieve many monotonous and unpleasant driving scenarios.
The two biggest challenges facing the future of AVs are technology and profitable business models.
In terms of technology, it’s critical that AV technology truly works. Michelle Avary, head of autonomous mobility at the World Economic Forum, says, “Really making sure that the technology is working in the areas of perception, which is vision— being able to identify objects and then understand how to move around them. That has yet to be solved.”
Avary also believes that the ongoing trade tensions between the United States and China could prevent firms from sharing geography-specific data sets and from operating outside their own countries. It may “actually stymie the growth of the industry,” since collaboration and data sharing among companies building AV technology would be squashed.
According to Avary, the business model for self-driving vehicles is also quite an undertaking. “We see some big divergence between the whole idea of the business model of the robo-taxi versus what we see in areas like commercial trucking, mining, and construction, where the business model case might be more readily made,” says Avery.
Compared to the so-called robo-taxis focusing on transporting people, using autonomous vehicles to transport goods on highways is a more lucrative business model. In the face of an impending shortage of truck drivers, self-driving trucks could lower the cost of shipping goods by eliminating human drivers. There are also lucrative opportunities for AV technologies to be utilized in mining and construction industries. For example, autonomous machines could dig trenches for laying oil pipelines.
Preparing for the Future
No matter the timetable, AV technologies are progressing and continue to promise improved driving safety. Be sure your organization is prepared with IEEE Guide to Autonomous Vehicle Technology. This seven-course program covers foundational and practical applications of autonomous, connected, and intelligent vehicle technologies.
Choudhury, Saheli Roy. (1 Jul 2019). Self-driving cars face two important challenges, says World Economic Forum executive. CNBC.
Kaslikowski, Adam. (30 Jun 2019). Everything you need to know about autonomous vehicles. Digital Trends.
Edelstein, Stephen. (10 Oct 2018). Cadillac Super Cruise beats Tesla Autopilot in Consumer Reports testing. Digital Trends.