Thursday, December 27, 2012

Exposure to Air Pollution

You may travel green, going by bike or ride a bus, but you may put yourself at possible health risks. The Atlantic Cities published an article on environmental exposure:
"The participants doing the most to reduce carbon emissions by cycling or riding the bus were the people who experienced the highest levels of exposure to pollutants."
Hints: Avoid cycling in busy streets plus stand away from the curb and wait behind the bus shelter.

Here is a link to the summary of the research done at UCSD:

Link to a paper by Adam Moore from Portland State University on "An Empirical Study of Particulate Matter Exposure for Passengers Waiting at Bus Stop Shelters in Portland, Oregon, USA."

Sunday, December 23, 2012

Expansion of Turkey's airlines

"Going head-to-head with Qatar and the United Arab Emirates (UAE), two countries that have placed aviation at the heart of their economic development and have seemingly bottomless pockets to fund their strategies, may seem risky. But Turkey has some advantages. Istanbul is within four hours' flying time of most of western Europe, the world's largest market for international travel, as well as being equidistant from much of the Middle East, North Africa and central Asia. This allows Pegasus and Turkish Airlines to deploy smaller narrow-body planes from their hubs, affording greater flexibility for route planning than is enjoyed by the widebody-operators of the Gulf. Turkey's size also gives it a strong home market, guaranteeing buoyant domestic demand."

Wednesday, December 19, 2012

Trip Length Distribution in Chicago

The average trip length (distance) in Chicago, regardless of the trip purpose and mode, is 7.9 miles. Following is the distribution of trip length in Chicago based on the 2007-08 Chicago Household Travel Survey data. There are many micro-trips (super-short trips) recorded in the database. Examples of these micro-trips are the short trips to buy coffee or lunch from a nearby store while at work. You may think of more examples if you just recall your own trips in the last few days.

Beside the super-short trips, the distribution is obviously has a long tail to the right. One interesting point is that 50% of the recorded trips are less than 5 miles.

Two comments:
  • Auto share for these trips (I mean trips of less than 5 miles which comprise 50% of all the trips in Chicago) is more than 60%. I think, 5 miles can be easily traveled with a bicycle. Just imagine for these trips, if people choose to bike instead of driving. This means that about 30% less vehicles will be moving around the city (with single occupancy assumption). 30% less vehicles in Chicago is roughly equivalent to 4 million vehicle-trips (Chicago has about 13 million vehicle-trips in a day [2007 estimation], predicted to be more than 30 millions in 2016).
  • How does trip length change over years? How is it correlated with land use and sprawl? If a city grows in size, does the average trip length increase? How is trip length distribution in Chicago different from other cities? These are some of the questions that I would like to look into in near future.
* UPDATE 1: The average trip distance in the Los Angeles metropolitan area is approximately 13 miles based on the data presented in a SCAG report (Figure 5-3, page 56).

Population density of the Los Angeles metropolitan area: 2,645 /sq. mi (1,025 /km²)
Population density of the Chicago metropolitan area: 1,318 /sq mi (509 /km²)

Although the population density of Chicago is lower than Los Angeles, trip distances are shorter in Chicago, perhaps because of a more mixed-use land development (requires further research).

* UPDATE 2: The average trip distance in the New York City is approximately 8 miles (about 12 miles for Manhattan) based on the data presented in a NYS DOT report (Table 63-b, page 6-12).

Monday, December 17, 2012

Mode Choice of Short Trips: The Absence of Biking

How do travelers behave when it comes to short trips? Here, I define short trips as trips less than 1 or 2 miles. The following graph is plotted based on the 2007-08 Chicago Household Travel Survey Data. For trips less than 1 mile, 84% of travelers choose to walk. However, there are still some people (11%) who drive. For trips less than 2-mile, the walking share drops significantly to 32% while driving share jumps to 61%. Biking and transit shares did not change considerably.

It is understandable that walking for distances longer than 1 mile is not something that many people can physically do. An alternative mode for a little bit longer distances (up to 2 or 3 miles) is biking which is apparently sadly absent in the choice set of many travelers in Chicago. A large portion of the 61% mode share of driving for less than 2-mile trips can be easily done with biking if sufficient biking infrastructure are built and (as importantly) people gets educated and encouraged to bike.

How long people walk and bike?

Walking and biking distance depends on many factors, obviously, such as age, trip purpose, weather, quality of built environment, personal characteristics, etc. Some studies have suggested that people tend to walk longer distances to transit stations. The following figures show the distribution of walking and biking distances  in Chicago based on the 2007-08 Chicago Household Travel Survey Data. Many outliers in the data are removed before plotting. Click on the figures to enlarge.

85% of walking trips are within the 0.75-mile range.
85% of biking trips are within the 3.5-mile range.

n (walking) = 10,673
n (biking) = 1,136

Here is a link to a nice article on walking distance to transit:

Wednesday, December 12, 2012

Bicycle Radar Sensor Installation on Dearborn St

Chicago Bicycle Program posted these photos on its Facebook wall, announcing the installation of a bicycle radar sensor on Dearborn St:

"Did you know that Dearborn has Bike Radar? Two sets of radar sensors were installed just north of Polk St to detect southbound bicycles as they approach the light. The sensors communicate wirelessly with the stoplight and once a bicycle is detected the signal will change to a green bike light!"

Here it is one of the newly installed bicycle signal in Chicago:

Monday, December 10, 2012

Chicago Bike 2015 Plan

"The Chicago Bike 2015 Plan has two overall goals:
  • To increase bicycle use, so that 5 percent of all trips less than five miles are by bicycle.
  • To reduce the number of bicycle injuries by 50 percent from current levels."
To support the City of Chicago's plan, I would like to see what the current mode split is for short trips. The mode shares for short trips can be obtained from the 2007 Chicago Household Travel Survey. Also, it would be beneficial to develop a simple mode choice model for short trips specifically to learn more about travelers behavior when trip distances are less than 5 miles. Good research opportunities exist in this area.

The picture shows a recently developed buffered bike lane on Dearborn Street.

Saturday, December 8, 2012

Air Cargo Market

According to the Boeing Long-Term Market Development study, the air cargo market has been growing with an average rate of 5.5% since 1980. The study also states that "Air cargo traffic growth, measured in revenue tonne-kilometers (RTK), is projected to average 5.2 percent over the next 20 years." There seems to be many opportunities of research and business in the air cargo. One topic that I am interested in is the air cargo network design. Similar to passenger air networks which benefit from hub-and-spoke structure, air cargo networks can do the same if demand remains growing. Today, large air cargo companies have already chosen a hub for their service. For example, Memphis, TN is the main of the FedEx Express. The question is where the optimal location for placing a hub for air cargo is? How different air cargo demand is from the air passenger demand? and so on.

Thursday, December 6, 2012

Journal of Choice Modeling joins Elsevier

This morning, I received the following email from Stephane Hess, the Editor in chief of the Journal of Choice Modeling, that the journal will be published by Elsevier from January 2013. This is good and bad news. It is good because it may strengthen the impact of the journal. On the other hand, it is bad because JOCM is no longer an open access journal. 

Dear colleagues,
It gives us great pleasure to announce that from January 2013, the Journal of Choice Modelling (JOCM) will be published by Elsevier. First published in 2008, JOCM has established a reputation for publishing high quality research by leading authors across the many different disciplines in which choice modelling is a key analytical technique. By working together with a world leading publisher, we will be able to further strengthen the impact of the journal. It is our hope that publication of JOCM alongside Elsevier’s outstanding collections of economics and transportation journals will bring this important work to the attention of an even wider audience.
From 2013, JOCM will publish four issues per year, which will be included in existing ScienceDirect collections, making them available to a very broad group of readers. Additionally, the current back catalogue consisting of the first five volumes of JOCM will be hosted on Elsevier’s ScienceDirect journals platform where they will continue to be available free of charge to any reader. JOCM will also be included in the SCOPUS database, the world’s largest abstract and citation database of peer-reviewed literature, further increasing the impact of articles published in the journal. Chris Pringle, Executive Publisher at Elsevier, says, “Since the Journal of Choice Modelling’s interests extend from transport into marketing, finance, environmental economics and beyond, its focus on choice modelling nicely complements our existing journals in both economics and transport, in a growth area which we wanted to represent more strongly in our list. JOCM and its editors have a track record of demonstrated success, and we are very happy to have the opportunity to add such a high quality journal to the Elsevier family.”
We trust you share our excitement at this important new development for JOCM, which should further strengthen interest from readers as well as authors. We are now in a transition process and until the journal obtains its dedicated section on the Elsevier journal system, the current website will remain in use.
The existing editorial team will continue to remain in charge, as follows:
Editors in chief
Stephane Hess, University of Leeds, United Kingdom
John Rose, University of Sydney, Australia
Associate editors:
Michel Bierlaire, Ecole Polytechnique Fédérale de Lausanne, Switzerland
Michiel Bliemer, University of Sydney, Australia
Juan de Dios Ortúzar, Pontificia Universidad Católica de Chile
Riccardo Scarpa, Queen’s University Belfast, United Kingdom
Kenneth Train, University of California, Berkeley, United States

Tuesday, December 4, 2012

Postdoctoral Position at Ryerson University

[Note: The following e-mail message is being shared with all members and friends of the Transportation Network Modeling committee (ADB30) of TRB.]
The Department of Civil Engineering in the Faculty of Engineering and Architectural Science at Ryerson University is now accepting applications for a Postdoctoral Fellow to help initiate a number of newly funded, cutting edge transportation research projects. The research involves developing adaptive network design models for urban logistics planning and operations—encompassing new public transport business models—to accommodate Big Data analytics and activity-based travel demand through the use of cyber-physical transport systems.

This is a minimum one year position with potential for extension. The start date of the position will be in the early spring of 2013. The candidate needs to have a PhD in an area related to transport systems, as the research will broadly cover one or more of the following areas: public transport, network optimization, urban logistics, activity-based micro-simulation, approximate dynamic programming, machine learning, transport economics, and/or cyber-physical transport systems. The candidate should have excellent communication skills and show an interest in supervising graduate students. Interdisciplinary collaboration with other growing research initiatives within Ryerson is expected (for example, DMZ:, CUE:

Interested applicants should send an email to Dr. Joseph Chow ( with the following package: a CV, contact information of three references, and a statement of research interests.

Transportation During and After Hurricane Sandy

A report describing the effects of Hurricane Sandy on New York City's transportation systems is published by the New York University Rudin Center for Transportation.

Monday, December 3, 2012

Walkable City

Jeff Speck recent work is a book on "Walkable City". Jeff is a city planner and architectural designer.

Kaid Benfield just published an article online reviewing Speck's book. While I agree with them for the most part, I would like to give a heads up to city planners and decision makers that they should not rush into making their cities walkable by cutting space from vehicles and giving it to peds and bikes. The menu-type list that Speck and Benfield have provided, need to be implemented in real world simultaneously and in long-term to be effective. A wonderful transit service in the middle of an inhabitant dessert is as bad as encouraging people to bike from 50-miles away suburbs to downtown to work in winter. We need to remind ourselves that the travel choice that we make everyday is sometimes a result of some of our own habits and in some cases a collective effect of problematic habits of the entire society. In other saying, I would probably bike to work when I see most of my fellows bike. Other than that, our choice of where to live and where to work, availability of good schools for kids, safe neighborhoods, and many other factors affect our choices of travel and residency.

What I would disagree is that "Traffic studies are bullshit." First of all, traffic doesn't mean just vehicular traffic. In fact, traffic is multi-modal. Second, saying "put cars in their place" give the wrong impression that cars are bad. In opposite, I would say cars are good because they give us the mobility that we want in a world (I mean mostly cities in the U.S.) where we have no other choice for the most time. How am I supposed to go shopping when the only store close to where I live is 10 miles away and is a big box? How am I supposed to go to work when my office is in downtown and my house is in a suburb? The point that I am trying to make is the higher importance of mixed-use development at this stage. The listed items in the Speck and Benfield's recipes need to be weighted differently and ordered depending on the current state of our cities. What city planners and policy makers need to know is that before calling off all the traffic studies, shutting down freeway lanes, and hiking up parking fees, they need to change the design of our cities and make sure that other options than just driving are available and not just available but desirable too. Then by making driving more expensive (or better to say less desirable, because it's not just about money, it's also habit and culture), people would choose to take transit, to walk or to bike.

Just to give you some interesting numbers: the population density in Amsterdam (within city limits) is about 3,745 people per sq km while the population density in Chicago (within city limits) is about 4,466 people/ sq km. What? Chicago has a higher population density? How come we do not bike as much as Dutch? Well, because other factors are in effect not just population density. Let's just imagine, we move all the residents of Amsterdam to Chicago and take Chicagoans to Amsterdam for a year. What would happen to travel patterns in those cities? I doubt that if it remains the same. So, it's not just the design or infrastructure or land-use, its also habits, inertia, and culture.

Sunday, December 2, 2012

Center for Urban Science and Progress (CUSP)

Center for Urban Science and Progress (CUSP) was founded in 2011-12 by New York University and NYU-Poly. Steven Koonin was appointed as the founding Director of the Center.

"CUSP simultaneously responds to two important challenges:
  • The very specific challenge issued by the Bloomberg Administration to create an applied science institute in New York that will make the city a world capital of science and technology, and lead to new jobs, and
  • the grand technical, intellectual, engineering, academic, and human challenges posed by a rapidly urbanizing world."

Saturday, December 1, 2012

Extreme Bus Bunching or Data Error?

I was tracking CTA Bus #147 at home to make sure that I would leave my apartment on time to catch the bus and that's what I came across. I wasn't at the stop to confirm whether this was an actual extreme case of bus bunching or just an error in data. Based on my personal experience, CTA route 147 gets bunched frequently, despite being an express service for the most part.

Apparently, this was not the first case of extreme bus bunching. A similar case has recently happened in Seattle and reported on a blog.

CMAP Congestion Pricing Plan

Chicago Metropolitan Agency for Planning (CMAP) is planning to implement congestion pricing in the Chicago region as part of the GO TO 2040 plan. The current congestion pricing plan of CMAP only focuses on express toll lanes. The selected freeway segments for dynamic tolling include new lanes on the I-90 Addams Tollway, the I-290 Eisenhower Expressway, and the I-55 Stevenson Expressway plus two new facilities, the Illinois Route 53 north extension and Illinois Route 120 bypass and the Elgin-O'Hare West Bypass.

To support the plan, CMAP has released a report quantifying "some of the benefits of congestion pricing, such as travel time savings, and also examining potential drawbacks, such as impacts on local streets and inequity among users." The report has reasonably studied the benefits and drawbacks of the plan. However, I think it has room for improvement. To determine more effective pricing plans, a more comprehensive study is required. Specifically, existing time-dependent travel times on the selected corridors need to be compared with the modeled travel times before and after pricing. Currently, the CMAP report only uses daily VMT for validation purposes. A future step to improve the analysis would be a time-series analysis of traffic flow and speed considering the "temporal spreading" effect when roads are dynamically priced. Temporal spreading refers to the change in travelers' departure time to avoid high congestion prices in the peak periods. Also, the reported changes in travel times on arterials and freeways seem (to me) to be over estimated. Note that no discussion on "induced demand" is provided.

Moreover, congestion pricing should not only be limited to express toll lanes. Cordon pricing, such as those that have been implemented in London or Singapore, should be considered as well. Downtown Chicago, an area roughly bounded by the Division St on North, Halsted St on West, Roosevelt St on South, and Lakeshore Drive on East, is a highly congested area during morning and afternoon peak periods which can potentially be priced. An alternative plan to mitigate congestion in downtown Chicago is metering the input rate into downtown based on the prevailing congestion level. A similar plan has been implemented in Zurich, Switzerland.

Friday, November 30, 2012

Funded PhD Positions Available in the Areas of E-Car and E-Bike Sharing Systems

Two funded PhD positions in the areas of "Electromobility and E-Car Sharing Systems" and "E-Bike and E-Bike Sharing Systems" are available at the Institute for Traffic, Transport and Regional Planning at the University of the Federal Armed Forces Munich, Germany. For more information visit:

Thursday, November 29, 2012

Transit Ridership in Chicago since 1981

I plotted this graph using the ridership data provided by the The Regional Transportation Authority (RTA). The graph shows the total system ridership including Pace, Metra, CTA rail, and CTA bus in terms of annual unlinked passenger trips in millions over years since 1981. Transit ridership in Chicago was declining (on average) from 1981 to 1995. Since 1995, transit ridership is picking up again but it is far from its historical peak in 1981 with more than 800 millions unlinked trips. There is also a small peak in 2008 which is perhaps associated with the increase in gas prices. Also, note that the increasing trend in system ridership comes mostly from the roughly constant increase in CTA rail and Metra usage. Ridership for Pace and CTA bus services remains almost constant since 1995.

The question is whether the ridership continue to increase? What can planning and transit agencies do to maintain this increasing trend? What are the factors affecting transit ridership in Chicago?

"Why Are States Passing Up Billions in Federal Transit Funds?"

"Why Are States Passing Up Billions in Federal Transit Funds?" is a recent article on the Atlantic Cities.

Kelly Clifton, Associate Professor at Portland State University, asks:
"Is it urban vs. rural or problems coming up with match? Or just the desire for more roads?"
I personally think it's a combination of all. States with more rural areas (fewer large cities) are less interested to flex funds to transit. Perhaps, transit agencies are not also strong enough (or maybe interested enough) to lobby and make the "flexing" happen.

"The underlying point, emphasized by Tanya Snyder at Streetsblog in her coverage of the new report, is that states can decide to allocate more money to transit "without any changes to national transportation law."

Also, I believe that the transit ridership is the U.S. does not increase unless people's travel behavior changes. Such change does not happen by only investing in transit infrastructure in an isolated manner. Other influential factors such as land-use need to change too. As long as population and job densities stay low, land-use remains unmixed (residential + commercial + institutional), and so on, it is unlikely to see a big shift in favor of transit, walking, and biking.

CMAP Modeling Research Opportunities (RFP)

Chicago Metropolitan Agency for Planning (CMAP) has recently announced two RFPs:

  • RFP 097 (Agent-based Economic Extension to Mesoscale Freight Model): The purpose of this RFP is to extend CMAP’s current freight modeling application to include a means for preparing future freight scenarios sensitive to distinct conditions at the global scale and explicit policy interventions from within the CMAP region.
  • RFP 096 (Network Micro-simulation Extension to Activity-based Travel Model): The purpose of this RFP is to extend application of an activity-based model of travel demand to include dynamic sensitivity to multi-modal network conditions.

The deadline for receipt of submissions in response to the RFP is 3:00 p.m., January 7, 2013.

Wednesday, November 28, 2012

I-55 Managed Lane Project

IDOT is initiating the I-55 Managed Lane Project in the Chicago area. According to the project website:
"The scope of work for this project is anticipated to include the addition of one lane in each direction within the existing median of I-55 needed to accommodate implementation of a managed lane, which could include Express Toll Lane (ETL), High Occupancy Vehicles (HOV) lane, High Occupancy Toll (HOT) lane, Congestion Priced lane, or other feasible managed lane configurations as determined to be appropriate for a projected 2040 travel demand."

Based on the findings in the literature and my personal research work, managed lanes need to be carefully planned and designed considering common issues such as "induced demand" and "empty lane syndrome". Adding capacity to a road, when not properly managed, can induce demand which can eventually result in the same level of congestion despite the added capacity. On the other hand, a HOV/HOT lane may also suffer from underutilization or "empty lane syndrome." Also from the operational point of view, an HOV lane works better if passes through one or more recurrent bottlenecks. If an HOV lane ends prior to a bottleneck, its capability to reduce congestion by smoothing the flow of HOVs may be harmed. One example of such not properly designed HOV lanes is the HOV lane on I-5 NB in Portland, Oregon which ends about 1 mile prior to a recurrent bottleneck. In this case, the bottleneck-induced congestion backs up and blocks the outflow of the HOV lane and as a result, a queue forms in the HOV lane, just like the adjacent general purpose lanes. 

Identifying the locations of the recurrent bottlenecks on I-55 should be done prior to any initial design of the HOV/HOT lane on this corridor in order to make sure that the beginning and end points of the managed lane is properly selected.

More Processors versus Run Times

A recent blog post by Marc-André Carle suggests that "using more processors does not necessarily lead to reduced run times on CPLEX."

Running 90 models with CPLEX 12.4, he concluded (not very strongly though) that "while increasing the number of processor cores results in quicker runs on average, but the effect on individual instances is quite difficult to predict."

Complex Systems and Networks

I have recently become very interested in complex systems. My question, which is not very hard to answer, is whether transportation networks and systems are complex? I would say yes, they are.

Following is a list of some of the reasons why a system might be considered to be complex, according to a recent paper, "Challenges in Complex Systems Science", published by Dirk Helbing and his colleagues:

  • Many heterogeneous interacting parts 
  • Interactions of autonomous agents

Transportation systems are complex because they consist of many heterogeneous interacting parts and agents. For example, a public transportation system in a city consists of bus routes, rail routes, park-and-ride facilities, and so on. Just imagine a Chicagoan commuter who lives in a suburb. Every morning she drives her car, picks up a colleague on her way to a Metra station (carpooling), parks her car in a designated park-and-ride facility and both ride Metra to downtown. Then from the Metra station in downtown, they transfer to a bus to get to their office. Throughout this commute trip, different heterogeneous parts and agents are indeed interacting. As another example, bus bunching (also train bunching) is a direct result of uncoordinated interactions between different agents (buses or trains) in a public transportation system.

Other characteristics of complex systems are:
  • Path-dependent dynamics
  • Complicated transition laws
  • Self-organization or collective shifts
  • Non-equilibrium dynamics
  • Adaptivity to changing environments
  • Multilevel dynamics and so on.

Some of my recent findings on network traffic science suggest that network traffic dynamics are path-dependent. In fact, in a paper published in TRR in 2012, "Exploring Properties of Network-wide Flow-Density Relations in a Freeway Network", we suggested a path-dependent characterization of hysteresis in network traffic. Hysteretic transitions in traffic, either in macro scale on a single facility or in network scale, can be considered as complicated transitions laws. Self-organizing behavior of pedestrian crowds and the existence of the Network Fundamental Diagram (NFD) of road networks, as a collective effect of interacting vehicles in a network, are another examples.

Networks: An Introduction by Mark Newman

Networks: An Introduction
By M. E. J. Newman
Hardback, 784 pages
Oxford University Press, March 2010
ISBN13: 9780199206650
ISBN10: 0199206651

"The study of networks, including computer networks, social networks, and biological networks, has attracted an enormous amount of interest in the last few years. The rise of the Internet and the wide availability of inexpensive computers have made it possible to gather and analyze network data on an unprecedented scale, and the development of new theoretical tools has allowed us to extract knowledge from networks of many different kinds. The study of networks is broadly interdisciplinary and central developments have occurred in many fields, including mathematics, physics, computer and information sciences, biology, and the social sciences. This book brings together for the first time the most important breakthroughs in each of these fields and presents them in a coherent fashion, highlighting the strong interconnections between work in different areas. Topics covered include the measurement and structure of networks, methods for analyzing network data, including methods developed in physics, statistics, and sociology, graph theory, computer algorithms, mathematical models of networks, including random graph models and generative models, and theories of dynamical processes taking place on networks."