GM EN-V Concept Car: 'Small, Electric, and Networked'.
Have you heard of the GM En-V? It's a concept electric vehicle, sort of a cross between a personal electric car and a cell phone, rolled out at the 2011 CES.
You know, I think that whether or not the actual vehicle makes it into production, the concepts introduced and demonstrated by GM's En-V are actually inevitable, and this is the most interesting concept vehicle I've seen yet.
Happily, I got a chance to chat by email the other day with Chris Borroni-Bird, General Motors Director of Advanced Technology Vehicle Concepts.
From Concept to Production
The first we saw of GM's En-V in the US, I think, was the 2011 Consumer Electronics Show in Vegas. Correct? Why are they going to consumer electronics shows rather than car shows?
We chose to demonstrate the EN-V at the Consumer Electronics Show this year because that venue provided the opportunity for driving demonstrations including autonomous operation.
Meaning they drive themselves....?
And currently autonomous driving with the EN-V primarily works outdoors because of the need for a GPS signal. During the 2010 Shanghai World Expo we demonstrated autonomous driving indoors through the use of dead reckoning, but that only works over short distances.
Since the EN-V makes extensive use of network communications technology we felt that the CES would be a good venue for a demonstration.
The EN-V concepts have also been on static display at other events including the 2011 North American International Auto Show in Detroit and the TED conference, and we've also done driving demonstrations at the New York Auto show and the Google I/O conference.
I like the idea that you can retrieve your car from anywhere, using your smart phone. With autonomous driving, is it possible to put a non-driver in the vehicle alone and send them to a destination?
One of best features of autonomous vehicles like EN-V is the ability to provide personal urban mobility for everyone. Parents could strap their children into an EN-V, program the destination and send them off to school while older people that can no longer drive could still easily get out of the house and live an active and independent life. Autonomous vehicles would also allow people with physical impairments such as blindness or paralysis to get around independently.
Would these vehicles require dedicated roadways, or are the onboard sensors adequate to prevent crashes in real life driving conditions (pedestrians, bikes, dogs, falling debris, 18-wheelers)?
No, ideally vehicles like this would operate on existing roads with other vehicles. The cost of a new independent infrastructure would probably be too high to justify.
Alternatively, some city centers in Europe have already decided to ban traditional automobiles and create car-free city centers and these could be good places to re-introduce an optimized form of personal mobility, such as EN-V, because weather conditions, long distances, human frailty are all reasons why cycling and walking are not always sufficient solutions.
Cities like Beijing, Shanghai and even Manhattan don't have any room to expand road infrastructure while demand for vehicles continues to climb. One of the best ways to increase personal mobility is to reduce the physical footprint of vehicles allowing more of them to fit on the existing roads as well as requiring less valuable real estate for parking).
While the collision avoidance technology on the EN-V today works for demonstrations, it must be made more robust to varying real-world conditions. Before we can rely totally on such systems, we'll have to prove that they are reliable and durable.
Loss of network communications can be tolerated for smartphones but could be a showstopper for V2V.
Even sensing technology needs to be proven out and chaotic city centers, even at low speed, is a very challenging environment for robust autonomous operation.
I heard that these vehicles would help to prevent pedestrian and bike fatalities. Can you explain a bit the technology that makes this possible?
The EN-V concepts use a combination of ultrasonic sensors and optical digital cameras to detect people and objects in their path. The images from the camera are processed to detect pedestrians and automatically slow down to avoid contact. Other sensors under consideration include radar and lidar systems.
There's autonomous and semi-autonomous driving possible with the En-V; is there non-autonomous driving possible, too? Is there manual override of the automatic systems in case of an emergency?
Sure, the EN-Vs can be operated in a fully manual mode, too.
So, has there been any testing done to see how these perform in real life driving conditions?
No, not yet. These current EN-Vs are concepts that were developed for the 2010 Shanghai world expo, and they were designed for operation on flat smooth surfaces. The concepts have no suspension and the low ground clearance limits the ability to go up and down grades, and they're not street legal since they don't meet any existing safety standards.
These are strictly proof of concept to demonstrate the possibilities of a small, networked autonomous electric vehicle.
Clearly, they don't need to be 2 wheeled but could have 3 or 4 and they could also be holding anywhere from 1 to 4 passengers so there is a lot of design freedom but any solution should be small, electric and networked, capable of manual and autonomous operation, may be private or shared ownership, and should help with integration with public transport.
So no crash testing?
No, they don't meet crash standards. We're actually intending that they don't crash. The idea would be that vehicles drive autonomously and do not crash into each other.
Any thoughts on how they would perform in the snow, or in windy conditions?
We don't have any real-world data under these conditions but with their low ground clearance they were not designed for extreme weather conditions as they are concepts which are not intended to be operated in those environments.
...or on hilly terrain, like San Francisco?
The current vehicle is not designed for hilly terrain and would not work in a place like San Francisco, but we're developing a next generation EN-V-type vehicle that is optimized for real-world operation. The goal is to begin some real world field trials in the next few years to demonstrate the viability of this type of vehicle.
The En-V in Emerging Markets
I understand that the En-V concept is at least 10 years from production. Is that true, and if so, is it also true in China and India, since China is hoping for a million electric cars on the road by 2015?
The basic technology that makes the EN-V concepts work today is pretty much off-the-shelf, but we'd have to make significant enhancements to it before this type of vehicle would be viable on the road.
The autonomous driving systems. GPS is critical to the autonomous operation of the EN-V but current commercial GPS is only accurate to about 3-5 meters and is often not reliable in urban areas with a lot of tall buildings.
Fully autonomous operation requires sub-1-meter positioning accuracy and more detailed map data. As more GPS satellites are launched in the coming years, precision and reliability will be improved but it is likely to be at least a decade before GPS is robust enough for this type of application.
Another critical element of making an EN-V-type vehicle viable is the collision avoidance technology. If a vehicle can actively avoid collisions, much of the crash safety structure and features like airbags found in contemporary vehicles can be eliminated which allows for dramatic reductions in weight and size. We will have to demonstrate the robustness of these accident avoidance measures before existing vehicle safety regulations are changed to allow the use of EN-V on the road.
Sensing is a near-term approach to enabling autonomous operation and is being developed by GM and others, such as Google. Vehicle to vehicle communications (V2V) is effectively a 6th sense although more information can be communicated than just the safety-critical. For example, location-based services and parking availability information could be communicated with vehicle to infrastructure (V2I). V2I can also complement sensing where sensors are not effective, such as 'seeing' around corners.
Are En-Vs (potentially) coming to those emerging markets first, and if so, where will they be built?
A small, inexpensive vehicle like the EN-V is ideally suited to crowded megacities in emerging markets like China and India but is also well suited to developed markets where congestion and parking are issues, such as European cities and some cities in the Americas. However, a suitable charging and communications infrastructure would need to be put in place wherever these vehicles are deployed. At this time General Motors has not announced any production plans for a vehicle of this type.
Are there plans to make them for US markets, and if so, where and when?
A vehicle like the EN-V could work just as well in U.S. cities as it does anywhere else. Again, at this time General Motors has not announced any production plans for a vehicle of this type.
Potential Applications for the En-V
Teen drivers. Are there any plans to marry the En-V to graduated licensing programs? It seems a natural fit for teen drivers, who are the most dangerous drivers on US roads, to have the option of driving a low-speed vehicle that only carries two people and is completely connected - could even offer a sort of electronic "fencing", I suppose, to keep kids within 20 miles of home or whatever. Thoughts?
Young drivers are just one of several segments that could benefit from this type of vehicle.
With it's autonomous capabilities enabling geo-fencing, a vehicle like EN-V could automatically limit where a young driver goes, how fast they go and prevent accidents. This same capability would make it attractive for car sharing programs.
Elderly and disabled drivers. Particularly with the autonomous drive feature, I can see these being viable as an alternative to losing your license when your vision becomes impaired or your reflexes too slow to drive safely. Can these vehicles be modified to accommodate the mobility-impaired, or to replace the mobility scooter?
Enabling people with limited mobility to maintain an independent and active lifestyle was one of the many factors that drove the development of EN-V. The skateboard chassis design of the EN-V places all of the drivetrain and energy storage hardware in the bottom of the vehicle and enables a wide variety of different body styles to be used on top including designs with wheelchair accessibility.
Thank you, Chris.
I'm particularly jazzed about the applications, because as a mom, I'm thinking about how many teenagers crash their cars every year - often with tragic consequences - and how preventable those crashes could be with this technology.
Please hurry with production on these, GM, before my little Gavin turns 16!