Driving forward: What’s the state of autonomous vehicles today?

Also in the U.S., Walmart is using autonomous cargo vans to deliver groceries in Arizona, while Pizza Hut is working with Toyota on a driverless electric delivery vehicle that even has a mobile kitchen in it to cook pizzas en route to your house.

Renault is one traditional car manufacturer making great strides in autonomous vehicles. Having been the company behind the Symbol in 2017, the first vehicle designed for mind-off automation, Renault recently announced it was conducting tests of on-demand car services in Paris and partnering with Waymo to investigate driverless transportation possibilities to and from Charles de Gaulle Airport. Renault’s Valerie Pecresse has said that the company is “investing 100 million euros in autonomous vehicle infrastructure and [we hope] that we will be able to bring this project to a successful conclusion for the Paris Olympic Games.”

Tesla is also making big steps forward in taking autonomy into mainstream use, both in terms of real world use cases and potential monetization of self-driving technologies. Tesla has supplied customers with more than 780,000 vehicles since launching, the majority of which arrive with pre-installed, self-driving capabilities available to users who purchase the requisite software. Tesla autonomous vehicles have logged huge levels of miles driven since their introduction, growing from 0.1 billion miles in May 2016 to an estimated 1.88 billion miles as of October 2019.

Ford is another manufacturer with deployments already in play, with self-driving vehicles being tested in Pittsburgh, Palo Alto, Miami, Washington D.C. and Detroit, with Austin, Texas joining them soon. Together with its partner Argo AI, Ford has plans to trial its fleet of self-driving cars in Austin with a view towards launching a wider-reaching autonomous taxi and delivery service in 2021.

Autonomous vehicles in first gear

There are a few things slowing autonomous vehicle deployments, ranging from regulation issues to business models to the technology itself. In terms of regulation for example, Waymo might have demonstrated a viable autonomous taxi service for places in California and wants to deploy it more widely as soon as possible, but the state’s government isn’t making it simple: California has said Waymo has to offer the service for free. Some industry analysts see this attitude as jeopardizing California’s reputation as the world leader in driverless technology.

GM subsidiary Cruise, which GM bought for $1 billion in 2016, has also experienced delays to its deployments in New York, after safety concerns kept the company from gaining regulatory approval from the state government.

Business models remain a challenge to deployments, too. Michelle Avary, Head of Autonomous Mobility at the World Economic Forum, has said, “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.” Car-as-a-service might be one workable way forward, with around half of car owners recently admitting they would consider giving up car ownership in the future if on-demand autonomous mobility solutions were available to them. But it’s something car manufacturers are still trying to work out.

Levels of automation

There is another barrier to autonomous vehicle deployments at the moment – levels of autonomous driving technology. These are guidelines issued by the Society of Automotive Engineers (SAE) to explain the various levels of autonomy, with Level 1, driver assistance, relating to computer assistance of simple driving functions like the cruise control or automated braking systems. Level 2 refers to partial automation, where the vehicle assists drivers with steering or acceleration, allowing drivers to disengage from some tasks. Level 3 concerns conditional automation, where the vehicle takes over some of the monitoring of the environment from the driver, using sensor technology like LiDAR. Level 4 is high automation, where much greater control has been handed to the vehicle, which is in charge of steering, braking, accelerating, monitoring the vehicle and roads, and also responding to events like deciding when to change lanes, turn or use signals.

Audi introduced Level 3 autonomous cars in 2018, while Ford stated its intention to go straight to Level 4, with a spokesperson saying previously, “We’re not going to ask the driver to instantaneously intervene – that’s not a fair proposition.” There is a Level 5 – full automation – in which the vehicle is in control of everything, and the proposition is that the human driver has to do nothing. This is not presently a realistic goal though, with Frost and Sullivan’s Connected and Autonomous Vehicles: Winning the Global Race to Market report predicting that Level 5 autonomous vehicles are unlikely to be introduced before 2025.

5G and autonomous vehicles

France is one country that is making big steps forward in supporting autonomous vehicles. March 2019 saw Renault demo a concept car that it intends to deploy as an electric robot taxi. Orange has an agreement to work with Renault on a connected vehicle and 5G project: the scheme will measure the capacity of 5G and how it can be used to enrich communications between vehicles and their environments, giving autonomous vehicles more information and enabling better traffic flow and improved road safety.

Orange is also working with Intel-owned Mobileye, a market leader in Advanced Driver Assistance Systems (ADAS), data collection and autonomous vehicle technology, to provide Internet of Things (IoT) connectivity for its Mobileye 8 Connect solution. Again, the goal of this initiative is to make roads safer and also give municipalities the data they need to monitor infrastructure and enable smart cities.

The 5G factor

5G mobile connectivity will be central to the future of autonomous, connected vehicles and essential if Level 5 fully-automated vehicles are to ever become a reality. With that in mind, Orange teamed up with PSA Group in 2017 to start testing 5G network technology for connected vehicle applications.

5G will enable IoT capabilities in autonomous cars: cars will be equipped with sensors that help make them smarter, and these sensors will generate massive amounts of data. Handling, processing and analyzing these huge amounts of data will need a much faster network than existing 4G technologies. Autonomous cars and their automated systems will need unprecedented data processing speeds if they are to be able to mimic the timing and speed of human reflexes. Furthermore, 5G will enable vehicle-to-vehicle (V2V) and vehicle-to-everything (V2X) connectivity in cars, while its low latency will also help make vehicles safer and more reliable.

A connected-vehicle future?

As it stands, there aren’t yet any companies offering fully autonomous car services on major roads, meaning the long-pictured image of sitting back and watching a movie in your car while your car drives itself is not here yet. But it’s in the works.

According to Counterpoint Research’s Internet of Things Tracker, the global market for connected cars will have grown 270% by 2022, with more than 125 million passenger cars with embedded connectivity set to ship worldwide between 2018 and 2022. Germany, France and the UK are all forecast to hit nearly 100% connected car penetration by 2020. Exciting times.

For more information, read how Orange and MobileEye are making connected vehicles and roads safer through data gathering and analysis.