11.S197 / 11.S944
Applied Urban Analytics: Modeling Pedestrian Activity in Cities

Permission of Instructor
Limited to 12 students

The course investigates the interaction between pedestrian activity, urban form and land-use patterns in relatively dense urban environments. Informed by recent literature on pedestrian mobility, behavior and biases, we take a hands-on view and learn how to operationalize and model pedestrian activity in built environments using software tools and analysis methods. Rather than engaging in comprehensive travel demand modeling across all modes, we use simplified, yet powerful and scalable network analysis methods that focus uniquely on pedestrians. Emphasis is placed on not only modeling or predicting pedestrian activity in given built settings, but also on analyzing and understanding how changes in the built environment—land use changes, density changes, connectivity changes—can affect pedestrian activity.  

The course is structured around three experiments:

  • Pedestrian accessibility analysis.
  • Trip distribution, facility patronage and critical route detection.
  • Pedestrian impact assessment triggered by land use or development changes.

Each experiment runs for four weeks, during which groups of participants are asked to tackle a real-world analysis exercise from beginning to end, starting with an introduction of theory and methods, followed by data collection and analysis and ending with a presentation of findings in class. By exposing class participants to different experimental set-ups that move from conceptualization and experimental design, to data collection, analysis, to the presentation and interpretation of findings, the course aims to prepare students for real-world walkability analysis work.

Each experiment is conducted in teams. A positive and constructive attitude for team-work is essential for a successful completion of the course.

There is no mid-term or final exam; each of the three experiments counts equally, distributing the workload throughout the semester. Class meetings introduce students to both relevant theory and software applications needed for each experiment.

It is recommended that participating students have taken at least one basic statistics course. Prior knowledge of Rhinoceros and ArcGIS are helpful too, though not required.

The course is very hands-on. We use multiple software platforms including Rhino, ArcGIS and Excel, along with Urban Network Analysis tools that are new and experimental. If you are willing to embrace learning by doing, you should experience enough to become a self-sufficient learner in urban analytics and pedestrian modeling, and might discover a whole new lens through which to study, plan and design built environments.