I’ve spent the vast majority of my career worrying about mobility. Optimizing traffic control, planning Intelligent Transportation Systems, and years of developing travel demand models all of which focus on measures of mobility. Everything we are trying to improve – speeds, delay, congestion, and intersection Level of Service – all are measures of mobility. And in every case, mobility is measuring how easily we can move around by car.
Mobility measurement isn't enough
In the fields of transportation engineering, planning, and modeling, it has been very well documented that as we add more capacity to roadways, you don’t relieve congestion in a 1-to-1 ratio. Inevitably, more vehicles appear filling up the capacity and increasing the total Vehicle Miles Traveled (VMT). We affectionately call this latent traffic demand. But this game of add capacity, increase VMT, doesn't slow us down. We keep trying to catch up by building more capacity and when money is short, we focus on operational improvements to squeeze more capacity out of the system. This vicious circle is easily justified as we show demands on our infrastructure growing faster than we can add more capacity. But what if we have the whole game wrong? What if adding capacity is what is growing the demands on the roadways?
This type of paradox has been a key theme this year. I am hearing it again and again as I am meeting customers around the world. They have reframed the question from "How do I move more cars?" to "Why are the cars there?"
For me, the key moment this year was a trip to visit Michael DuRoss at the Delaware DOT. I went there wanting to know more about their new Delmarva Multi-State Freight Model. Michael was kind enough to entertain my discussion, but he was very keen to share with me the work they have been doing with accessibility modeling in Cube. You can read more about the study here, but check out the following short highlights reel of our discussion.
The list below hardly captures the extent of their work, but I walked away with a fresh understanding of the following:
1) They are able to explain more trip making behavior by considering a household's accessibility score in addition to using demographic attributes alone.
2) It wasn't until they scored individual properties, instead of blocks or neighborhoods, that they were able to explain the details of walking trips in an area. Now using Cube and ArcGIS, they can easily model any area in the state at a very detailed level.
3) Using this dynamic modeling process, they test localized changes, the new trail connections below, which have very real impacts on vehicular travel as well as pedestrian, bike, and transit trips.
In my mind, their findings intuitively made sense and the Activity Modeling proponents have been pushing these things for some time. Using myself as an example, when we need bread — I can run over to a local store (5 min trip?) and pick it up. If what I need is not available at the local market, it’ll go on a shopping list, and I’ll stop at a larger grocery or big box store on my way home from work. I get it. The more accessible a location is, measured by the number of activities I could satisfy locally, the more likely I will make shorter, local trips. The shorter the trip, the more likely I am to walk. And where this really gets interesting is that not only am I reducing my vehicle miles traveled, but I am making more trips overall and these trips are generally to businesses which improve the local economy. I also find that the less time I spend in navigating large parking lots and traffic in general, my quality of life improves.
Where do we go from here?
Accessibility is primarily a question of land-use and then secondarily connectivity. Our traditional measures of roadway capacity and level of service have little to do with it. So what can I do about these things?
For starters, we acknowledge it, and — as the Delaware DOT has done — we start measuring it. In addition to measuring only speeds, congestion, and level of service, it’s time to start looking at Access Scores, as well. Measure peoples’ ability to reach valued destinations. As we start measuring and documenting accessibility, I’m confident we’ll find many ways to improve and optimize accessibility, just as we have done for all the measures of mobility.
Sugar Access - an ArcGIS extension for measuring Accessibility
Professional users of Cube have been building and calculating accessibility measures, but often it isn’t trivial. Sometimes, the biggest challenge can be just gathering and setting up the necessary detailed networks and datasets. To facilitate anyone’s analysis of accessibility, Citilabs has created a very simple ArcGIS extension called Sugar Access that comes with a complete local dataset — all setup and ready to go. Check out the webpage and recent technical webinar.
I’d love to hear more about if you and your agency are looking at accessibility and how you are incorporating it into your planning, engineering, and modeling today.
For more information
Mr. Martimo is a technical manager who was instrumental in integrating traffic simulation and ArcGIS into Cube and led the design and development of Cube Cloud. He also served as Vice President of Research & Development for AirSage, a pioneer of emerging technology in wireless signaling, location analytics, and big data processing. Mr. Martimo is now responsible for customer engagement and the international business groups at Citilabs. Matthew holds a bachelor’s degree in transportation engineering from North Dakota State University.
firstname.lastname@example.org | P: (+1) 404 671 9223