How can transportation care for the elderly? Will there ever be a convenient solution for that distance from your house to a public transit station? How much do our own behaviors influence AI?

Dr. Susan Shaheen is a pioneer in transportation and she was thinking about connected and autonomous mobility long before it was front page news. Whenever smart people come together to talk about the future of transportation, the sharing economy, and the effect of it all on our environment Susan is among them, often leading the conversation around research she has or is doing as the co-director of the Transportation Sustainability Research Center (TSRC) at the University of California (UC), Berkeley. We stole a little bit of her time at this years Transportation Research Board meeting in Washington, D.C. and picked her brain about what really is around the corner for transportation.


“Late-Night Transportation: How Two Public Agencies Are Filling Service Gaps Through Mobility on Demand” by Susan Shaheen on Move Forward



Regina: (00:00) From GRIDSMART Technologies. I’m Regina Hopper. Welcome to POLICYSMART.

Regina: (00:18) Welcome to this edition of POLICYSMART. This is a person that I have wanted to talk to now for over the last year because every time I see smart people getting together, she’s leading them. {laughs} Susan Shaheen, adjunct professor at UC Berkeley. You are actually a pioneer in the area of intelligent transportation and new mobility. So thank you for taking time to talk with us.

Susan: (00:46) Oh my pleasure. Delighted to spend time with you.

Regina: (00:49) So back in the 1990s you started researching what we now call mobility on demand. Um, you were a pioneer like I said, because you understood I think at that time that you had environmental issues that could be covered, equity issues, that could be covered, access issues that could be covered. So just tell me how did you get there before everyone else did?

Susan: (01:15) That’s a great question and I think about it a lot cause I do get asked this question and I think what happened for me is that I came into transportation wanting to focus and do something different. And to me technology was a solution that could possibly help us address some of these policy barriers that we just couldn’t get to, you know, like pricing by uh, a carbon tax or road pricing, things that are just somewhat politically and intractable. Right? And so how could we use technology to produce something that was more efficient, potentially more attractive and could open up a whole new way of traveling for people. And I had this hypothesis that if you took these fixed costs of auto ownership and translated them into variable costs, which is essentially what mobility on demand or mobility as service is trying to do, is to commodify each of those individual trips. If we could provide a superior way of traveling, maybe people would be willing to pay more per trip, get rid of that car and then potentially start making very different decisions about their overall modal choice.

Susan: (02:34) Perhaps taking public transit, more car pooling, walking, cycling, and the adventure I’ve been on for the last 20 years –

Regina: (02:42) Gosh, isn’t that crazy?

Susan: (02:43) Yeah, UNTANGLING each and every new shared mode, how its impacts are different. And now even looking at how their impacts are different based on spatial and temporal dimensions.

Regina: (02:57) Now you’re going to have to, now this is where the, Yup. This is where the professor comes in. So for those who are listening, when you say spatial and temporal, give us a little bit of like 101 on that one.

Susan: (03:08) Yeah. So when a lot of these services first started, you’d see them in the urban core. So that’s your spacial dimension.

Regina: (03:16) Oh, okay. Mostly in the large cities.

Susan: (03:18) Right. And also when we talk about his temporal dimension, are those trips being taken on the evenings and weekends or are they being taken at peak when we have a lot of roadway congestion? So it’s really the wearing when that we’re honing in on more and more layer by layer, study by study, mode by mode. And what I have to tell you is that the first one that we spend a lot of time on car sharing systems like Zipcar are different than even one-way car sharing like a car to go and like sharing’s different ride sourcing or Uber, Lyft type models are different. And every single one is doing different things, attracting different types of riders in some cases. And this is also part of the question of equity is who’s using these and are some able to provide different levels of accessibility than others and why?

Regina: (04:18) From a business perspective, how long does it take to run these things through a system to see if they’re going to work or not? I mean, so I was surprised. I was in Dallas like right before the holidays and those little scooters were everywhere. I really hadn’t even noticed them in DC. But boy in Dallas, they were like on every corner. And I thought to myself, Gosh, this change is so quick. Even when you’re in the industry, this changes so quickly. So from a business perspective, how long does it take before we understand if that mobility type is going to work?

Susan: (04:51) I think we’re still learning, to be honest with you. These are evolutions and they’re tweaking their models. You, you might not be aware of it. In some cases you will be more aware because it will show up in your pricing. But in other ways it’s just they’re providing a new type of service, like an Uber pool or Lyft shared ride in which you can reduce your costs but maybe increase your travel time by sharing that ride with someone else. And so with each of these tweaks, does that help more in that location? Does it help to grow or do they need to scale back?

Susan: (05:25) Focus on different features later. So it’s very challenging to understand. Now the oldest of these models that we’ve studied is car sharing and we know a lot –

Regina: (05:36) Like the Zipcar?

Susan: (05:37) Like the zipcar. So we know a lot more about the marginal costs of providing these services and where they work and where they don’t. They tend to be phenomenon of the urban core or dense environments like a college university campus or an employment site. Right? But then you’ve got these on demand, one way types of services which include car sharing, micro transit services, which is like a larger occupancy vehicle, like a shuttle. And then you have, in addition to that, the Ubers and Lyfts, right? And so trying to figure out economic viability is a big question. And another thing that’s happening right on top of all of this and you’re busy, you’re bringing in and the business perspective of this, a little bit of venture capital can go a long way in terms of providing those scooters all across that landscape in Dallas.

Susan: (06:31) But we don’t necessarily know if it’s going to be a fad or if it’s going to be a long term pattern. Now we’re pretty excited about the idea of active transportation, but will that particular business model of a standing east scooter be one that has longevity, will people once they sort of get used to them just go buy their own?

Regina: (06:53) Well, I wondered about that.

Susan: (06:54) Right?

Regina: (06:55) You know, it’s like cause, you could manufacture that and just own it, right?

Susan: (06:58) That’s correct.

Regina: (06:58) Move on.

Susan: (06:59) Right. Just like a lot of people own their own electric bike or um, standard bike. So, you know, do the dynamics in people’s travel patterns change? Is, it is a really significant question that I think a lot of us are trying to answer. And ultimately going back to the roots of my research for my doctorate, how is this affecting car ownership and use?

Regina: (07:23) Yeah, I was going to ask you about that with the Ubers and Lyfts. I was going to climb back up.

Susan: (07:27) Good, good, good, good.

Regina: (07:29) Go ahead. Talk about it because I mean, I know you see the General Motors and the Fords and the Audis and name them all who are trying to do these business deals now because they know that maybe car owners, I’m at my bond, my car anymore.

Susan: (07:42) Right. Oh I know.

Regina: (07:44) So talk a little bit about that.

Susan: (07:45) Yeah, so our research over the last 15 years indicates that there are people who used all of these types of services who choose to sell an existing car or suppressed or postpone a car purchase. So these are real effects. However, they are very difficult to quantify.

Regina: (08:06) Because I guess you have to be able to assess, Hi, I’m Regina, I didn’t buy a new car because I can get an Uber easily based upon the fact that I live and work around the DC area. All those things have to come together, right?

Susan: (08:20) That’s correct.

Regina: (08:20) Okay.

Susan: (08:21) But the question I’m going to ask you is, so you really think that Uber and Lyft caused you to do this? Okay. You’ve stated this in your, your question response. Now I need to say how much were you traveling on that private car before you sold it? Can you quantify that for me?

Regina: (08:42) Yeah. Everywhere.

Susan: (08:44) How much has the arrival of an innovative service caused you to take that action? Or was it because you shifted jobs or housing location?

Regina: (08:55) Yeah, that’s interesting.

Susan: (08:56) So the reason this is a really exciting area of study is that if we can unlock this, perhaps we can start to unlock a lot of things, get people taking transit more, get people walking and cycling more, get people using all of that money that’s dedicated to those fixed capital costs on the outset on something else like education, housing, better food quality.

Regina: (09:20) You know, leaving the equity issues aside, and to me, one of the things I love about ITS intelligent transportation is that equity seems to be so much of a part of everyone’s conversation. But if you set that aside just for now, you know, you wonder like that last mile to me has always been the most difficult even personally. Right?

Susan: (09:42) Sure

Regina: (09:42) So I don’t take the subway because, and I have to walk too far or you know, whatever that last mile thing is, is this, are these new systems of transportation really focusing on that last mile effort?

Susan: (09:58) There’s a lot of focus on first mile, last mile applications because we see sort of system or transportation network effects buy in that and coming from a place like the United States, we have a lot of people living more in suburban areas, edge cities, places where making that connection to public transit can be challenging and difficult and may be enough of a barrier that causes them to revert or to just take up car ownership and use. So there’s a lot of focus on that particular application here in the US there is around the world as well. But I would say our particular land uses have created these challenges of first mile last mile in a more pronounced way of saving then a dense urban area of, of Europe. I’m sure you’ve traveled, right?

Regina: (10:47) Yeah. Oh yeah.

Susan: (10:48) It’s not, it’s not that challenging to get to and from the public transit station, right. It’s like when you live in the suburbs of Virginia or Maryland, that can be a challenge. But I wanted to tell you that there’s other applications that we’re looking at that are pretty exciting beyond first mile. Last mile is can we fill transportation gaps like late night transportation. We just did a, a blog on this, uh, which I released late last week.

Regina: (11:16) Oh, well, we will want to link it to this. We want to make sure so people can find it.

Susan: (11:19) I’d like people to think about this is um, there’s people who work late at night and they don’t have a lot of good safe quality transportation options other than a private car. So filling that gap when public transit doesn’t have headways or people feel unsafe, even if there are infrequent headways is a really big challenge. And so can shared mobility options, fill some of those gaps, making it easier for people who need to travel late at night. And let’s face it, a lot of people don’t work traditional hours. And a need and need options. The other thing that’s really exciting is this idea of downsizing some of those buses that run around and suburban areas and they’re empty, but they’re taking a lot of our public transit dollars. Can we downsize them into shuttles or can we provide on demand mobility options instead of running those empty buses, which isn’t good for our planet either. So that’s an area that’s really exciting is possible replacement of existing parts that are not functioning well and then just filling gaps where there was no transit option at all available.

Susan: (12:34) And, and how could that open up mobility and accessibility, particularly for people who didn’t have it. You know, there’s studies that indicate that Uber and Lyft type services are going into some locations where taxis wouldn’t go.

Regina: (12:49) Yeah, yeah, no, that’s true. And even smaller communities. I remember when when Uber and Lyft for started, now we’re out of Knoxville, GRIDSMART’s out of Knoxville, you couldn’t find an Uber, right? It was too small of a community, even with the university’s segment there. But now it’s almost as if you can go just about anywhere and find an uber or lyft, right.

Susan: (13:09) Your wait time’s just are going to differ based on those densities. Right. And, and some of your costs might also be a reflection of that. But uh, some of my very early research on the, on the concept of carpooling indicated to me that if you were below pretty consistently a 15 or 10 minute wait time, people who were willing to accommodate that and, and so people’s expectations and the city core may be different on wait time then they are, as you get into the suburban areas or edge times, you just plan for it. It was the inconsistency. Right?

Regina: (13:44) Right.

Susan: (13:44) The lack of reliability that stood for so long as a barrier to two people sharing rides more frequently. Yeah. I remember not wanting to catch to catch a bus in New York. Right. Because I didn’t know how much time to add in, so I wouldn’t be late. So let’s talk a little bit about how people adapt to this because one of the things that we find in autonomous and connected and trying to explain this to people, my 86 year old father now is all about autonomous and connected vehicles. But a lot of people were like, I don’t want that. I want, I either want my car or I want my, my mobility on demand, my Uber, or my Lyft, but I want a human being. So when you talk about opening up the technological aspects of transportation, what’s the thing you worry most about in adaptation or acceptability from the general public?

Susan: (14:36) Uh, well I think perceptions around safety of automated vehicles is one of the most important things that we work on is helping people understand what the safety benefits of an automated vehicle are. I think when we talk about making these automated vehicles more efficient, potentially having them pulled, we have to also think about perceptions of safety of being in a vehicle with no driver on board.

Regina: (15:05) Right.

Susan: (15:06) So, and then when we talk about elder care, right. Which is near and dear to my heart. And that was actually another area I was going to tell you about, right. That’s opening up and really exciting in terms of market opportunity or use case is paratransit is getting your elderly dad to and from his medical appointments point to point and maybe ultimately door to door, but with uh, people from the silent generation or even the baby boomer generation, they may not have a mindset that is as accepting of technology.

Regina: (15:40) Yeah.

Susan: (15:40) Automated or intelligent. And so how do we address those issues while also providing them with potentially outstanding opportunities to get around so that they don’t suffer from social isolationism or lack of access to quality food or health care.

Regina: (15:58) Yeah. It affords for the elderly. This could be a brand new world, right. I mean, even if it’s just something as simple as companionship. I know being able to, to, to get with their friends in a way. I mean, my father goes back and forth now about should I drive, should I not drive? That kind of thing. And, and it’s hard, you know, because then you start making decisions in a different way because you start feeling isolated. We talk with someone yesterday who was talking about the equity issues of you addressed them, people who work at night, people who don’t have the economic means maybe to just whenever they want an Uber and Lyft to pay 20 or 30 bucks to do something. The elderly. So as you’re walking now into 2019 and we’re here at TRB and Susan, by the way, for the listener, Susan’s like on every panel, it’s like she can’t, she said I can’t be on everything, but she, she’s so in demand. What do you think going into 2019 is the issue around technology-based transportation? If you could wean it down to one or two?

Susan: (16:58) Well, one of the things I’m really curious about is how artificial intelligence and machine learning could have positive outcomes or in contrast, unintended negative outcomes and how we really need to get ahead of this and to shape this so that we’re not training the technology to have some of these biases that unfortunately we have as a society. For example, towards racial profiling, right? And so as we’re training these cars to drive and be driverless and automated or autonomous, how are they learning? Well they’re learning from us. They’re learning from our behaviors and they’re getting better. That’s the way they’re smoother. They drive a little bit more like a person now than they did in the late 1990s when I rode in them. So they are learning. But if we’re going to provide, for example, a shared automated vehicle, are they learning from the existing drivers and what choices they may make in terms of the kinds of passengers they pick up, the wait times, those individuals may experience and the locations that they may or may not be willing to enter.

Regina: (18:22) Oh that’s, you know what? I have to tell you, I have not even really thought about an AI being learning from our own biases. That’s really interesting. It was somebody’s got to teach the machine, right? I mean –

Susan: (18:36) We are teaching the machines.

Regina: (18:40) That is fascinating. So I wasn’t going to ask you about this. But so, so basically then the researchers then that are pushing an AI development who they are as important as the theology of AI itself, right?

Susan: (19:01) Yes. So how about an accident if you have more in common, are you, is your automated vehicle going to save you versus a person who’s a bike or pedestrian and doesn’t have automation protecting it? Are you going to be the priority because you’re on board? Does that ultimately cause us to want to leave our vehicles because we don’t feel safe when we’re not paying for protection? These are extremely important ethical questions. Moral questions.

Regina: (19:38) Yeah. We could talk a long time just about that, couldn’t we? That’s really interesting.

Susan: (19:42) The other thing I wanted to tell you Regina, um, while this is really fascinating and I wanted to bring it up to you because I think a lot about it, like when I’m laying in bed, oh my gosh, like what are all the implications of this?

Regina: (19:53) Right, right.

Susan: (19:53) But I think the other thing for 2019 that we really do need to stay focused on is as these lines blur between the public sector and the private sector, what is the role of government and all of this? How can we continue to prioritize and protect the public good. And that’s everything from social equity as we’ve been discussing to the environment. These are really big issues of our times.

Regina: (20:21) Yeah. And we see, I mean, I either have the ability and honor to be in Washington or I, it’s cursed or something. I don’t know, but you know, we don’t see the federal government being able to keep up with the technological trends that they try. It’s not the, that the bureaucracy doesn’t try. It does, but it because of how it’s instituted at camp. So what we see are states and localities and the private industry kind of going ahead and moving this thing out. Right. But it’s almost like piece part. So it may work this way in California or it might work that way in Arizona. Or if I’m driving down to the border of Iowa from the neighboring state, I might not be able to get in, does that concern you at all? With regard to the deployment of intelligent systems?

Susan: (21:03) Well I think standards are a big issue and how they’re deployed and the kinds of data that are being requested is a big deal. And what we definitely see in intelligent transportation systems is a largely unequal terrain. And I think this is going to make it challenging for us to ensure a certain level of quality or standard across your experiences between states in particular, but also for those manufacturers or services to understand how to provide the data because those requests may be different. And that’s why we have seen levels of standardization in many areas like uh, the national highway traffic safety administration, right? Having national levels of standards. So again, I think we’ve gotten multiple levels of governance and we’ve got a lot of questions about what is the role of government, not just at the national level, but at the state and local level. And a lot of these services, the rubber hits the road. If it’d be a bike tire or a scooter tire or a car tire, it hits the road on those local streets.

Regina: (22:17) Yep.

Susan: (22:17) And the local curve. So it’s all these different levels of governance that, that I think we really have to stay connected to. And the third issue that I think we really need to be thinking about in 2019 not that we weren’t before, is what’s the impact of these services on public transportation?

Regina: (22:36) Hm. You mean because it becomes so individualized that the mass may go away and just try?

Susan: (22:43) And it may be easier to get from point a to point B in a single vehicle than it is to make multimodal trips. So what does that mean for the public transit infrastructure? How do we maintain the public interest and support for those types of services and how can we create a more holistic operationally efficient system? And as you know, a lot of the public transit ridership numbers are currently in decline. Right? I’m not suggesting that’s entirely due to shared mobility services. And in fact, I would argue that it’s not, there’s economic issues. There’s failing systems that were not built to keep going as long as they’ve kept going. Yeah. There’s rising fairs, right, right, right. There’s economic issues in the background. Uh, so there’s crime and safety considerations and perceptions. So there’s a lot going on. So it’s not just shared mobility services, but like all ultimately, what is the role of public transit? Are we going to keep it? Are we going to it? Are we going to grow it? And how do we work with the public sector, private sector on that?

Regina: (23:51) So AI, public policy, transit, teaching these machines the right way. Oh my gosh, you must tell it good. You must have a whiteboard where you take the issue and just start bifurcating it down. Right? That’s just, there’s so much there. And you’ve been on the cutting edge of all of it. So we’re just so fortunate to have this a little bit of time.

Susan: (24:14) I’m fortunate to get a chance to, to hang out and talk to you.

Regina: (24:17) Well thank you.

Susan: (24:17) You and I have not been able to do that.

Regina: (24:19) I know and I know we want to more. As a matter of fact, as we get into 2019 we’d love to like catch up with you more as you see sort of where some of this public policy is going.

Susan: (24:28) We’d love to give you a first bite when we released some of our latest research.

Regina: (24:32) That’d be great. We’d love it.

Susan: (24:33) We would unpack it and help people understand it, at least from our perspective. That’s great.

Regina: (24:38) Thank you so much, Susan.

Susan: (24:39) My pleasure.

Regina: (24:41) Okay, so thanks so much for listening and participating in the POLICYSMART podcast. You can download the POLICYSMART app on the Apple App Store and Google Play and you can listen to the podcast on Apple Podcast, Google Play, and now on Spotify. And also please leave a review to help people like you, those interested in the next generation of mobility and intelligent transportation find us.


Recent Blogs

Stay informed and connected with industry insights, company news and personal stories

Protect Your Community with GRIDSMART Protect

Safety is the most fundamental need in your intersections. But existing in-ground detection and approach-based systems do not do enough, especially for vulnerable road users. That’s why Cubic Transportation Systems is introducing GRIDSMART Protect....

GRIDSMART Releases Version 19.12

Dear Valued Partners, Earlier this month, we shared a preview of GRIDSMART System Software Version 19.12 in a webinar that many of you attended. We're happy to share that 19.12 is now in general release and available for download. We encourage you to update all of...

Featured Event


After much deliberation, and with much remorse, GRIDSMART has decided to cancel INTERSECT this year. The safety of our employees, distributors and customers is of utmost importance.