 What do you think about the traffic here in DC? It's terrible. Very bad. Very bad. We have a really bad situation here in Washington. It's hard to drive. Very hard to drive. And do you think traffic's getting worse in Washington DC or better over time? I think it's getting worse. Yeah. Been out here for some time, I think it's getting worse. What's the worst you've ever been stuck in traffic here? Worst? Eight hours. Eight hours? Yeah. What did you do for eight hours? I tried to get home. One of the best ways to get somebody started in a conversation here in Washington DC is to start talking about traffic. Now everybody seems to have an opinion and it's mostly pretty negative. Hi, I'm Carl Wunderlich. I live and work here in the Washington DC metropolitan area. I know firsthand how difficult and stressful travel in this area can be. Our lives are increasingly schedule-driven. Not just work-related appointments, but in our personal lives as well. Daycare pickups, getting to the airport, are just being on time for a soccer practice. The bottom line is that we frequently need to be on time, or else the fragile chain of connected work and personal commitments begins to unravel. At the same time that our lives become more scheduled, travel has become more congested, more unpredictable. In order to cope with the stress of having to be on time in an unpredictable system, we find ourselves leaving earlier, mapping out alternative routes, and of course complaining a lot about traffic. Someone that's heard more than his share of complaints over the years is here with us today. He's the former Maryland State Highway Administrator, Hal Kassoff. Hal, in your tenure, I'll bet you've gotten more than an earful of complaints about traffic in the Washington DC metropolitan area. There's no question about it. The worst part of the state in terms of traffic was right here in the Washington suburbs and what we found was that it just affected people's outlook on life. I mean where they lived and where they worked and access to health care and jobs and child care, it affected the business community in terms of where they would locate. We'd have folks coming in from other parts of the country really trying to figure out an optimal place to be to avoid the worst of traffic congestion. It was a major topic of conversation. Sure, sure. Now, I know that provision of travel information is an important part of Maryland's approach to congestion. Now, how did you get started in travel information? Well, that's an interesting story. We had a newly elected governor, and he had campaigned that he was going to improve the worst trip in Maryland, which was the trip from the Baltimore, Washington area to our ocean resorts in Ocean City, Maryland. A trip that would normally take perhaps two to three hours at most in the summertime during the peaks would take five, six, even seven hours. Wow. He said, well, I've got some bad news because we're going to improve it by this summer. That was six months away, and we were desperate. So we turned to a softer approach, and the softer approach turned out to be reaching out to our customers, providing them with travel information. And boy, did that make a difference. So what do you think are the best ways of talking about the benefits of travel information services? Well, you know, we have our old set of tools that we looked at the system as transportation planners and traffic engineers and said volume capacity ratios or absolute travel times. And really, those were not the critical issues we found to our customers. So what we found is we have to start measuring in ways that respond to their needs. For example, how reliable is the system? Do I have to allow a cushion of time? How predictable is the traffic jam going to be? How much time do I have to budget? Do I have to leave earlier? So we need performance measures that resonate in those ways. Well, Hal, it sounds like the things you're talking about are very different than the way that civil engineers and traffic engineers typically talk about improvements to a roadway system. No question it's different. And the challenge is going to be to define measures that our civil engineering community, our traffic engineering community, can use to respond to customers in ways that relate to their issues. Sure, sure. And I think that in the next 10 or 15 minutes, Hal, we're going to take you through some new techniques that we have of using archive data to answer just those sorts of questions, try and measure those things like on-time reliability and predictability and travel budget in ways that we haven't been able to before. And so I think if you'll stay with us for the next 15 minutes or so, I'll walk you through some of that analysis. That sounds great. Can you mind if I take some notes? That'll be great. Do you find that the traffic reports on the radio help at all? Do they help? No, I pretty much do what I'm going to do anyway. So when you hear about a report of congestion, do you drive right at the congestion? Yeah, if everyone else takes their advice and goes off, I have the road to myself and it's usually cleared by the time I get there. So how's traffic here in Washington, D.C.? Terrible. Is today particularly a bad day? It's about the same as usual, but it's a little heavier where I normally go. So you had to bail out of your... this is not your normal route here? No, it's not. I'm just taking a different route to see how it goes. And how's it working out so far? Pretty bad. I'm talking to you. Traveler information. Traditionally, we get it by looking up at an overhead message board or by tuning in to a traffic report on the radio. But today, advanced traveler information systems are really changing the way that information is getting to us as travelers. And in fact, these new systems go way beyond a simple traffic report or a congestion-ahead message. For travelers on our urban roadway systems, these new advanced services provide personalized traffic reports directly to the traveler. Now the traveler can access the reports from a computer at home or at work, over the phone by dialing 511, or alternatively, by having a service, contact the traveler themselves when conditions are unusual. And this contact can be done in a variety of media, by cell phone, by pager, or by fax. These services seem to offer help in the struggle to manage the unpredictable reality of urban travel. But how effective are these services in saving time and reducing stress? Further, can these savings be accurately measured? In order to illustrate the benefits of using an advanced travel information service, let's take a look at a long, undesirable trip from the Washington metropolitan area. Actually, it's my trip from work in Laurel, Maryland to my home in Dale City, Virginia. And by showing you this trip, I hope to illustrate how not only we can talk about the benefits of using advanced travel information services, but also how we can measure those benefits as well. It's four o'clock here in Laurel, Maryland, which means it's time for me to go home from work. My destination is Dale City, Virginia, that's approximately 65 miles away. Unfortunately, it's across the entire Washington metropolitan area. And since I leave here about four o'clock every day, that means that I'm pretty much going into the teeth of rush hour traffic here. This commute is fairly unpleasant, but the one good thing about it is that I've got three different ways I can go. My first option is to follow I-95, crossing the Potomac River then heading south to Dale City. My second option is to head onto I-495, crossing the Potomac at the American Legion Bridge, then heading south on I-95 at Springfield. My third but least reliable option is to take the BW Parkway, cutting through the district to cross the Potomac at the Woodrow Wilson Bridge. Here's what I'm going to do. I'm going to take I-95 today because on average, it turns out to be the best for me. Traffic moves along well until I hit just before the Woodrow Wilson Bridge. Wow, one, two, three, four, five car collisions. Five drivers standing there talking to one another. The rest of us stuck. Continued heavy traffic beyond the bridge means that by five o'clock I've only reached Springfield, Virginia. Not good. This is not good. Ow! Well, that didn't work out very well at all because of the traffic now. I'm 25 minutes late for my daycare constraint here in Dale City. And for me, that means a pretty serious fine. In fact, when I get home, I get fine for every minute I'm late after five o'clock. And today, 25 minutes late, that means the fine I pay is actually more than I pay for daycare in the entire week. So because of the traffic, this turned out to be a pretty terrible day, a very expensive day for me in terms of commuting. I know exactly what I would have done differently, but this is not a good commute at all. Clearly, my plan to chance it and leave at four o'clock did not turn out to be a good decision. What could I have done differently? In order to answer this question, we employ a new technique based on historical records of travel times around the area. From these records, we can recreate what would have happened to alter egos of mine who made different decisions Let's take a look at an experiment which deals with today's expensive failed commute from Laurel to Dale City. First, I'll dramatize what would have happened if I had arranged my life so that I might leave earlier despite the fact that my boss prefers that I stay until four o'clock. What if instead of leaving at four o'clock, I left work at 3.30 instead? Now, this means I'll have to get to work earlier, wake up earlier, but by leaving at 3.30, I should be pretty confident that I'll be late very frequently. From experience, I know that at this time of the day, going along I-495 over the American region bridge is on average a better route than going on I-95. Now we're in Virginia, this is a place where typically there can be some serious problems, but today, on those distances there aren't any problems at all. I'm just driving along here, speed limit, whatever. Now by looking at the historical record it seems to be lighter than normal on this particular day, for this particular route. The result is very little delay and a very early arrival in Hale City. Well, here it is. It's 4.20. I'm in Hale City. Plenty early. A little bit too early, I guess. I've got 40 minutes now that I can do something with, I guess, but since I had no idea that I was going to be so early, in some ways I feel like I've wasted 40 minutes of time here. I'm paying for childcare that I'm not really using, but given that I had no idea I was going to get it, I guess I'll just go in and get my daughter and go on from there. Being 40 minutes early isn't so terrific, but I guess it's better than being light. Our trip statistics show that by leaving earlier, I would have been able to realize an improved but still imperfect outcome. Now, let's consider a case in an advanced traveler information system to contact me about unusual conditions and advise me on when to leave and which route to take. So, can we get the specs on Monday? I think Tuesday may be a little bit more reasonable. Sure, Tuesday is fine. Do you have any other questions on the material? No, I think I'm good. Okay, great. Well, it looks like there's a problem on my commute. That's my service calling in to let me know. In fact, I need to take another route, and as well, I need to leave. I need to leave. It's 3.45 now. I need to be gone by 3.50 if I expect to be in Dale City by 5 o'clock. My service lets me know where the problems are, and today suggests that I go over the American Legion Bridge instead of my normal route. So, there's nothing really here that I can't put off until a little bit later. Nothing is more important than getting back to Dale City at 5. I need to clean up and head out. The service has taken me off of my normal route. And again, by looking at the historical record of travel time, I would experience very little delay until I reached the Tyson's Corner area. I've run into some traffic here, but it's just slow. I'm going between 25 and 30 miles an hour. So, I know that this is a little bit slower than I'd like to go, but it's still we're still making progress and I feel confident that the service is taking me the right way and that there's no major accident just ahead here. Congestion breaks up and the rest of the ride turns out to have little delay. The result is that I arrive in Dale City at 4.50. Well, this is great. I'm going to get here about 10 minutes early. I'll take that. Okay. Well, there's a lot of traffic there. Nothing serious. And here it is. We've got here 10 minutes early. So, that's a pretty good service, I guess, today. It worked out fine. Let me know when I needed to leave and I've arrived with 10 minutes to spare. That's great. Using the service turns out to be particularly useful today. Clearly, my results using the service although it is instructive to compare across these outcomes, I still might want to know what the best possible strategy was for that day, the strategy that would get me to Dale City precisely on time. Again, by referencing the historical record of travel time, we can establish what might have happened if I had a completely accurate forecast of congestion conditions. Next, I'll be dramatizing a character that represents a hypothetical benchmark rather than a strategy I can actually adopt. I mean, hey, I'm Mr. On Time. I'll be there at five. Don't worry. All right, I'll see you then. Bye. How can I be so sure I'm going to be in Dale City at five o'clock? Well, I've got a view on exactly where the congestion is today. In fact, every day. So, I can decide to leave exactly at the right time that's going to get me to Dale City at five o'clock. I know which route to take. I know when to leave. And in fact, today, I see it's four o' three, let's go. Keep in mind that by calculating this optimal behavior, we create a benchmark for comparison with our other commuting strategies and their associated alter egos. Clearly, it does not represent a behavior that I or anyone else can adopt, whether we use advanced traveler information systems or not. The man says, right on time, guess what time it is? Five o'clock. Exactly five o'clock. Why? Because I knew exactly where the traffic was going to be, exactly where the congestion was going to be and exactly what route to take. So, here it is, five o'clock, Dale City. Can't do any better than that. We've seen how the use of information could be useful on a particular day on my trip from Laurel to Dale City. But what about over an entire year of commuting? Well, since we can go back to the records and construct any day in that period, we can take a look at on-time reliability over the entire period. The result is that if I choose to chance it every day and leave it four o'clock, I'm going to be late 50% of the time. But by leaving early, I can make that much better 97% of the time. If I use a service, I get fairly close to that 92% on-time reliability. And we can compare those results against the benchmark, the optimal character who has essentially a view of everything that's going to happen on his trip before he starts. The penalty for leaving early every day at 3.30 is shown in a metric like travel budget. Here we see that by leaving at 3.30, my leave early alter ego has to allocate 90 minutes to his travel every day. And it can be reduced either by chanceing it or by using a service. And you can see those numbers in the low 70s and compare that against our optimal benchmark who has a 66-minute travel budget. To summarize the impact of travel information on my trip from Laurel to Del City, the first thing we can say is that by using information I'm on-time quite reliably. Further, compared to the leave early alter ego who budgets a lot of extra time in, there's a 13% reduction in travel budget which turns out to be more than 80 hours an entire year. Overall, I can characterize the trip as being more predictable which means lower stress than if I try and chance it wasted time than if I leave early. We might ask, is Laurel to Del City the only trip in the DC area that travel information can be effective? And the answer is no. We took a look at a very representative sample of many different trips long and short across the DC area. And the results in terms of on-time reliability and our other metrics are actually quite similar. Altogether the key impact of travel information was more predictable travel. More predictable travel means less stress by traveling in a system because we know when things are bad and when they're good. And as well a reduced travel budget compared to behaviors that accommodate unpredictability by merely scheduling more time into an already scheduled life. Well Hal, now that you've seen the video what do you think are the key points? Well the first key point is that from our customer's perspective traffic congestion causes real stress in their lives. And it creates an obligation on the part of transportation people to do two things that I can think of. One is how can we provide them with the kind of information about congestion on a real-time basis that puts them in control of the situation. That enables them to select from options and make choices. That's number one. And number two is how can we measure the effectiveness of doing that. What are the benefits? We're talking about making investments in advanced technologies intelligent transportation systems travel information systems how do we justify that in terms of benefits. But benefits using measures that respond to our customers and what they value not necessarily the old measures that we like to use in the old days. Well Hal I think you hit the nail right on the head I think that the critical point is that if we just provide information to the traveler they're best suited to figure out what to do with that for their own benefit. They can make very powerful decisions. They can decide not to take a trip they can decide to take another route or leave a little bit earlier. It's up to them. It empowers them and I think that really is the benefit of advanced travel information systems. Absolutely.