While much attention is placed on the autonomous car in terms of safety and driving comfort, rarely does the discussion turn to its effect on powertrain architecture.
With all aspects of driving taken out of the hands of a human operator, what will be required of the engine and its associated drivetrain and transmission? The answer, says Dean Tomazic, Chief Technology Officer at FEV North America, will depend on whether the vehicle drives on an autonomous-only basis, or whether the driver can swap between conventional and automated driving.
Going fully autonomous
According to the scale of automation devised by Society of Automotive Engineers (SAE), a Level 5 vehicle has no requirement for a human operator to drive the vehicle manually, and will typically be used in a ride-share environment, or as a personal shuttle. In this case, the would-be driver will most likely use travel time as an opportunity to catch up on work, browse the web or simply relax. This means that harsh acceleration and braking is completely unnecessary in practically all scenarios, and here, the aim of the game is comfort.
Robo-cabs will have a completely different set of requirements to a vehicle that you and I currently drive. With an autonomous-only EV, for example, why would you need a 300kWh motor when 150kWh would be fine?
“With respect to the powertrain, the requirements for an autonomous-only vehicle will be simplified,” explains Tomazic. “Robo-cabs will have a completely different set of requirements to a vehicle that you and I currently drive. With an autonomous-only EV, for example, why would you need a 300kWh motor when 150kWh would be fine?”
But the same can also be said for an automated vehicle of Level 4 and below, which will be able to transfer between human and computer-driven control. In most cases, these vehicles will be privately-owned by those who still want to drive in certain situations, and do not want to relinquish their travel needs to a computer.
“Think about a vehicle that in the morning can drive you to work while you drink a coffee and check email, but then on the weekend allows you to go for a drive in the countryside and have some fun,” continues Tomazic. “The requirements are completely different in terms of driving dynamics; in the morning, you want a mellow experience so you don’t spill your coffee, but then the power and torque requirements are vastly different if you want to drive sportily.”
Automation will also have an impact on the lifecycle of a powertrain, and engineers will face something of a Catch-22 situation. In many cases, autonomous cars will be far better utilised than a traditional vehicle through the rise of on-demand mobility services. The powertrain will be subject to less aggressive driving – a point in terms of maintenance – but it will also clock up far more mileage.
Think about a vehicle that in the morning can drive you to work while you drink a coffee and check email, but then on the weekend allows you to go for a drive in the countryside and have some fun. The requirements are completely different in terms of driving dynamics
Autonomous ride-share services will have hundreds of vehicles in operation within a city, which may run 24 hours a day, seven days a week – a stark contrast to human working conditions today.
In May 2016, New York City authorities announced proposals to cut cab driver shifts to no longer than 12 hours, due to road safety concerns, and many drivers said to be ‘driving drowsy’. In July 2017, one Uber driver in Utah reportedly worked a 20-hour shift behind the wheel to maximise profits, and according to the European Transport Safety Council, no professional driver should drive continuously for more than two hours without “at least a 15-minute break”, although it says this is a “rule of thumb.”
Driverless vehicles will suffer no such issues, and will ultimately be able to operate continuously, pending mechanical failure. “With autonomous city driving through a ride-share or robo-cab model, the stress on components in the vehicle will not be as high as it is now,” agrees Tomazic. “From an acceleration perspective, the time-to-torque will also look different because it has to be a smooth ride. Passengers may be talking or reading email, and they don’t want to be jerked around. The requirements of the powertrain there are not as aggressive,” he affirms.
With autonomous city driving through a ride-share or robo-cab model, the stress on components will be reduced. From an acceleration perspective, the time-to-torque will also look different because it has to be a smooth ride
Indeed, demand in terms of acceleration and braking will not be as high as a privately-owned vehicle outside of the city, and this is likely to extend the lifetime of the engine, drivetrain and its associated components. However, that is not to say that these vehicles will run constantly for a duration of ten years before being replaced or repaired, points out Tomazic.
Uber, for example, requires that its vehicles were manufactured no earlier than 2010 in order to operate in New York City, and future vehicles may well be refreshed even more frequently. In future, Tomazic suggests that ride-share vehicles may operate for a maximum period of four-years before simply having the powertrain replaced or updated. “There is an inherent benefit [of autonomous driving] from a powertrain perspective as we know it,” he affirms.
Don’t forget the ICE
Writing for Greenstreetsoftware.info in July 2017, Lisa Jerram, Principal Research Analyst at Navigant Research, noted a likely “shift to electric vehicles in these automated shared mobility services”. But despite this, Tomazic points out that it is “definitely possible” for vehicles of Level 4 and below to use traditional ICE propulsion, and debunks the myth that automation could spell the end for diesel and gasoline cars.
We are looking at a wide variation of powertrain technologies that we will see in the field. It depends on who we are talking about, where people live and what they want to do with the vehicle
“ICEs will be part of the mix,” he says. “It is not only battery electric vehicles (BEVs) that will see autonomous driving capability.” For example, rural dwellers may want to commute into the city, and would be likely to do so in autonomous mode. However, the same consumer may wish to travel out of town – from Detroit to Chicago for instance. “These drivers might be limited from a range perspective with an EV, so an ICE may be required here,” suggests Tomazic.
As for automated fuel-cell vehicles (FCVs), don’t hold your breath. Today, a lack of infrastructure to support these vehicles outside of well-invested regions such as California will prove a particular sticking point. “An FCV with autonomous drive capability is probably not so high on the list of the OEMs right now,” he remarks. “FCVs are really the minority now, and even by 2035 could still only account for just 1% of all sales. Within the next ten years, I do not see them playing a major role.”
The underlying trend is that there will be no single answer to the powertrain as vehicles approach full automation. Powertrain architecture will vary depending on a range of factors, and will include both traditional combustion engines and electrified variants. “We are looking at a wide variation of powertrain technologies that we will see in the field,” Tomazic concludes. “It depends on who we are talking about, where people live and what they want to do with the vehicle.”
This article appeared in the Q3 2017 issue of Automotive Megatrends Magazine.