Skip to content
Who's in the Video
William J. Mitchell and the members of the MIT Smart Cities research group are creating innovative ways to change how we live in urban areas through, in part, the application[…]
Sign up for the Smarter Faster newsletter
A weekly newsletter featuring the biggest ideas from the smartest people

The U.S. can learn from European bicycle sharing programs and their lack of sophisticated solutions to system balancing.

Question: What are the components to Mobility on Demand?

Bill Mitchell: If you look at urban personal mobility, there are several components to the system. One component is the vehicle. What’s the vehicle like, is it an automobile, is it a bicycle, is it a scooter, is it some other kind of device? And so part of what we’ve been doing is reinventing the vehicle, sort of looking at small, light-weight automobiles and other kinds of vehicles as well. 

Another component of the system is the energy supply system, how to get energy to these things. And so the traditional way of doing this is distribute gasoline around the city. What we’ve been looking at, of course, is electrical recharging infrastructure.

A third part of it is, how do you gain access to vehicles when and where you need them, and what do you do with them when you’re not using them? And these were enormous problems. And our cities are choked with parking; it occupies an enormous amount of space that could be put to much better use, much more human and effective use for other purposes. It’s not always easy to get access to the vehicle that you need to go somewhere. You’re going to have to walk a long way often to get parking, or you can’t get a cab, or all of these kinds of things.

So, in response to this last issue that I’ve mentioned, this issue of how do you get access to vehicles and where do you put them, we drew up this concept we call Mobility on Demand because really in principle, it’s a very simple idea, but a very profound difference from the way we do it with private automobiles. But the idea is you have stacks of vehicles located around the city at conveniently spaced locations, when you want to go somewhere, you simply walk to a nearby stack, pick up a vehicle, drive up to a stack drop off point near where you want to go and drop it off. So it’s one-way rental, and the total time for the trip is the time to walk to a pick up point, do the transaction of picking up the vehicle, actually travel to the drop off point, and then walk to your final destination from there. We can demonstrate that we get with this sort of system that we can get much better door-to-door times under realistic conditions very often than you can get with a private automobile. And let’s suppose it – advantage of the private automobile, it’s there when and where you need it. That’s something of a myth actually because you have to go and dig it out of parking and then you don’t have a parking space, and so on. So we get great door-to-door times I believe out of this, our simulation suggests this to us.

We get very high utilization rate out of the vehicles in this sort of system, the vehicles are not sitting around doing nothing. And we occupy a minimal amount of urban real estate to do this. So, it’s a system that has a tremendous number of advantages if you can really make it work.

Now a fundamental technical issue with Mobility on Demand Systems, is how do you keep the system balance because, if you just think about all this in very rough intuitive terms, you can imagine easily that after a while, all of your vehicles are over here, but all of the demand is over here, so how do you keep the system balanced. And there are various ways to think about – the way that we do it and I think a very innovative way of doing it is with dynamic pricing incentives. So, on the fly, we create price incentives to return vehicles to locations that are not crowded that need vehicles and we create some incentives to return them to locations that already have plenty of vehicles. And similarly with pick up, we make it a bargain to pick up vehicles at locations where vehicles are plentiful and we make it a bit more expensive to pick up vehicles where they are scarce.

So, there is elasticity in the demand for mobility. I mean, we have a certain amount of choice about when we make trips and exactly where we make trips sometimes too. So we take advantage of this elasticity. We put the price incentives in there and I think we can show that we can very effectively balance this kind of system using price incentives. 

Question: What can we learn from bike sharing programs?

Bill Mitchell: The nearest thing to large scale Mobility on Demand Systems that exist are these European bike-sharing systems that do operate in exactly the way I described. They’re not electrical systems, they’re not automobiles, but they do operate – you pick up a bicycle and you ride it where you want to go and you drop it off, and they do have significant problems with system balancing. They really have not solved that problem very effectively. So you do find in those systems that sometimes you come to a pick up point and there’s no bicycle there, or you come to drop off. It’s even more frustrating you come to drop off and there’s nowhere to drop the bike off and it’s really annoying. And you have to ride off somewhere into the distance to actually park your bike and get a bus back to where you need to – this is obviously extremely frustrating. 

So, they have that problem. And then they have – mostly the way they do the system balancing is by picking up bicycles and throwing them on trucks and moving them back to where they need to go. At least I think this is the way they do it. This is what it looks like. I haven’t studied it carefully.

So, those systems suffer from not having a sophisticated effective solution to system balancing. Now, this is a prime example of why you need intelligence and the whole Smart Cities idea that intelligence is the key, I think, to sustainability and social equity in cities in the future because the dynamic pricing kind of structure that I’m talking about, or that I talked about a moment ago, depends on electronically keeping track of where the vehicles are, having sophisticated information processing so you dynamically adjust the prices and running an optimization algorithm. In many ways, it’s sort of something like, let’s say Google, which is a big complex system that in the end depends on some sophisticated algorithms to deliver to people the services they need, and this is very similar.

But mostly people have tended to think in the past about mobility, it’s like a matter of providing the hardware and providing access to the vehicles and physically moving people around. I would argue that in the future, it’s much more about sophisticated management algorithms and making the best use of the resources that you have through information technology.

Recorded on January 21, 2010


Related
The integration of artificial intelligence into public health could have revolutionary implications for the global south—if only it can get online.