Hands off the wheel!
A look at what the future holds for autonomous vehicles

Who’s going to drive you home?
It’s 8am and time for the morning commute. You get in your car, select ‘office’ on the sat nav, then sit back to catch up on the latest Netflix drama as your car reverses off your driveway and drives you to work. This is the future many envision with autonomous vehicles (AVs). But how far off is this reality? IET Fellow Naresh Ravuri, a Senior Engineering Manager at Manga, breaks down the future of self-driving cars, trains, planes, boats, and more.
An ageing idea
“This technology is not new. In aerospace, we use it for travel monitoring and object detection with radar and lidar, and we’ve had autopilot and advanced driver assistance systems for some time,” explains Naresh. The concept of AVs dates back almost a century to the first radio-controlled cars. In 2004, the US military offered a $1million prize for a self-driving car, though no one succeeded and in 2009, Google launched a self-driving car project for California- a service like Uber, but without a driver.
The problems to overcome
Current AV trials are in pre-geofenced areas, allowing the technology to map out an area and drive only on pre-trained roads. “This allows AVs to drive millions of miles, knowing an area very well. It began in California and is now in Austin, Phoenix, and soon Washington DC and Miami,” says Naresh. While there’s demand, setting up AVs is expensive due to the array of sensors and detection systems required.
What will become available sooner?
Fully autonomous vehicles may be far off, but various levels of autonomy will benefit the market soon. There are five levels, with the highest being fully autonomous. Level three provides a self-driving option where drivers still monitor the car. “Many of these vehicles are being developed in Germany and will hit the roads in the next few years. They will come with a disclaimer that if the vehicle cannot operate, the driver must take control,” Naresh explains. However, the challenge is ensuring drivers are ready to take control immediately if needed. “Someone may be sat at the wheel eating, drinking, chatting, reading a book, watching a movie – if you’re moving at speed and the car needs you to immediately take control, the two to three seconds it can take to do that can be an awfully long time in terms of avoiding a crash. We’re trying to integrate autonomous technology with humans, and the two don’t easily synchronise.”
It’s a challenge figuring out what the failsafe is when the driver is not responding. This is where driver monitoring systems within vehicles play a vital role in ensuring they are attentive before this handover may be required. “There are still many pitfalls around the technology, especially regarding its interface with human behaviour. That handshake between humans and technology is complicated and proves a challenge. It must predict what humans are going to do next.”
Neither manufacturers nor insurance companies want to be responsible for any crashes involving such vehicles when a driver fails to respond to a system request to take control, while regulations around the fast-moving technology are still being defined. Such cars would likely be in the higher price bracket as they become available, as early electric cars (EVs) were, though they would then have a system that would be fit for over a decade.
Do we need personal cars?
“At present, the concept of AVs is for commercial vehicles and mobility support rather than personal use,” says Naresh. In the long term, this may reduce the need for personal vehicles. “In a day I drive to the office, and I drive home, and the car just sits there when it’s not in use. Maybe commercial AVs can’t replace all home vehicles, but they could certainly remove the need for homes having more than one. The reduced operating expenses of not needing to maintain your own car would be a huge attraction.”
On the seas and in the skies
“Autonomous aeroplanes are still some way off. Autopilot is available outside of takeoff and landing, but fully autonomous flight is not imminent. We’ll have cars first,” Naresh says. In shipping, AVs are not likely soon due to the size and potential impact of accidents. “These are such big vehicles that if anything was to go wrong then the results would be huge. With planes and trains you’re looking at vehicles with large amounts of passengers travelling at very high speeds; a crash in such a scenario is disastrous, not like a rear-end prang at traffic lights you may have in a car. We do already have autonomous trains for some freight operations and short journeys, so this is likely to continue; things like taking the content away from mining sites, or connecting passengers around airport terminals, for example. I do expect to see autonomous trucks on the road soon, likely even before cars. There’s a big push from the commercial sector to get these operational.”
On your driveway?
The next few years are likely to see the wider rollout of the early stages of AVs, with level two becoming standard and level three being made widely available for personal vehicles. It may not be as simple as choosing the destination and leaving your car to it, but technology to considerably reduce the driver input needed is not far off. “The technology still needs a lot of improvement, so with so much work to be done there’s unlikely to be any job losses soon. Will the situation be the same in 20 years? Maybe not. There will still be tech eruptions though where people need to react to the latest innovation being delivered, but people may still need their own car for flexibility and mobility. Perhaps in 20 or 30 years teenagers won’t need to learn to drive at all though…”
The future may not be here yet, but it’s also not far off.