Final Project

The Final Project provides an opportunity to apply the skills and concepts developed throughout the course to a real-world urban planning challenge. In groups of 3–4, you will formulate a research question, identify relevant datasets, conduct in-depth analyses, and communicate your findings through impactful visualizations and interpretation.

You may choose any topic relevant to urban analytics (see the feasible and infeasible project examples on this page), or you may choose one of the topics offered through the MTC-Mentored Project Track.

You must use Python for all analysis. No other software or languages (e.g., ArcGIS Pro, QGIS, R) are allowed.

To support different project orientations, the final project offers two tracks, which share the same analytical foundation but differ in emphasis.

Project Tracks

Want more hands-on Python? Choose Track B. Both tracks write Python and both are graded on whether it runs cleanly. The difference is how far the analysis goes: Track A spends more of its effort on interpretation and visual design, Track B on a deeper and more rigorous analysis — larger or messier data, more demanding methods, and results you can defend.

Track A: Python for Data Interpretation & Visual Reasoning

For projects centered on interpreting urban data through maps and visual analysis. You still write and run Python; less of the grade rests on how the code is organized.

Deliverables

  • ArcGIS StoryMap (or equivalent)
  • 4+ standalone interactive HTML maps generated with Python
  • A PDF report

Emphasis

  • Visual reasoning and interpretability
  • Thoughtful use of spatial methods, robust cleaning process
  • Problem definition, story telling and visualization design judgment

Track B: Python for Data Analytics & Reproducibility

For projects that lean more hands-on with Python, where you want the engineering to count. Shorter written reports in interim phases.

Deliverables

  • GitHub repository + GitHub Pages webpage
  • 2+ standalone interactive and animated HTML maps generated with Python (integrated in the webpage)

Emphasis

  • Reproducible, modular workflows
  • Handling large or more complex datasets
  • Robust analytical methods

Opportunity Strong projects may be developed into workshop submissions, conference posters, or journal publications.

Important

Both tracks:

  • Use Python for data acquisition, cleaning, analysis, and visualization
  • Work with real urban datasets
  • Are evaluated using track-specific criteria aligned with their emphasis

Three Phases:

To ensure steady progress and tailored guidance, the project is structured into 3 phases. Detailed submission requirements for each phase are posted on this site as the phase approaches.

0. Group Formation

Deadline: Sep. 9th, 12:00 pm

Choose a team of 3-4 and fill this form out by the deadline.

1. Proposal and Lightning Talk Videos

Submission Deadline: Sep. 21st - 11:59 pm

This step ensures you are on the right track. The plan you outline here is a starting point and will likely evolve as you progress. You will document and justify any changes during the midterm submission phase.

2. Midterm

Submission Deadline: Nov. 2nd - 12:00 pm

There is no midterm exam. The midterm is a checkpoint on your project progress: clean data, a working workflow, and early results. The work does not need to be polished. See the Midterm page for the requirements, the submission checklist, and the Gradescope link.

3. Final Project Submission

Submission Deadline: Dec. 15th - 11:59 pm

The final is your project as a public, self-contained piece of work: the analysis finished, the maps interactive, and the whole thing readable by someone who was not in this class. The format depends on your track. Track A publishes an ArcGIS StoryMap, with the interactive HTML maps you generated in Python embedded in it. Track B publishes a GitHub Pages website, backed by the repository that produced it. Both are presented at a public exhibition during finals week, so the work needs to stand on its own without you narrating it. Start assembling the final artifact well before December: moving a finished analysis into a StoryMap or a webpage always takes longer than groups expect.

See Past Examples for projects from previous semesters in both formats.

Full requirements and the rubric for the final submission will be posted later in the semester, closer to the deadline. What is above is enough to plan around; the details will follow.

Examples of Feasible and Unfeasible Projects

Feasible Projects:

These are some examples of feasible projects, but you are encouraged to explore beyond them.

  1. Mapping Environmental Justice Disparities: Identifying Spatial Inequities in Exposure and Access
    • Example: Do school districts in Alameda County with lower median household incomes have lower levels of tree canopy coverage? (CP101 - Fall 2025)
  2. Investigating Food Deserts in City X: Spatial Correlates and Socioeconomic Impacts
  3. Unveiling Mobility Patterns: Analyzing Bike-Sharing Data to Enhance Accessibility
  4. Mobility patterns and access
    • Example: How does transit-based job accessibility change between the AM Peak (7:30–8:30) and Late Night (22:00–23:00) in 2018-2022 period, and do Equity Priority Community tracts experience larger losses than non-EPC tracts? (CP101 - Fall 2025)
    • Example: How did BART ridership collapse and incomplete recovery (2019-2024) differentially impact Berkeley’s three stations (Downtown Berkeley, North Berkeley, Ashby) across varying income levels and multimodal connectivity, and where did the missing regional riders go? (CP101 - Fall 2025)
  5. Urban Noise and Built Form: Examining Spatial Noise Distributions in Relation to Building Configurations
  6. Optimizing Public Transit: Data Cleaning and Visualization to Detect System Inefficiencies
  7. Crime and Urban Green Spaces: Evaluating Hotspot Patterns in Relation to Local Parks in City X
  8. Decoding 311 Complaints: Spatial Patterns and Predictors of Citizen-Reported Urban Issues in City X

Unfeasible Projects

These projects are either too broad, require advanced techniques beyond the course scope, or rely on inaccessible data:

  1. Predicting Urban Sprawl for the Next 10 Years
  2. Real-Time Traffic Monitoring Using CCTV Feeds
  3. Training a deep learning model from scratch to detect pedestrians in videos.
  4. Large-scale satellite image analysis requiring significant computational resources.
  5. Building a complex VR environment representing urban planning scenarios.
  6. Developing novel algorithms requiring extensive mathematical derivation.
  7. Simulating Climate Change Impact on Coastal Cities
  8. Proposing a new accessibility/walkability/… index

Remember, the goal is to showcase your skills in using Python for urban data analytics, geospatial analysis, and visualization.

MTC Mentorship Program Research Questions

1. Regional Housing Needs Allocation (RHNA)

Question: What methodologies should ABAG consider when allocating the share of Regional Housing Needs Determination (RHND) to each local government in the region for RHNA Cycle 7?

2. Defining Equity Priority Communities (EPCs)

Question: What methodology should MTC/ABAG use to define Equity Priority Communities (EPCs)?

3. SB 79 Transit-Oriented Development and Upzoning

Question: Compare some number of SB 79-affected counties, create maps and charts that show differences in TOD capacity, transit quality, and expected housing yield.

4. Daytime Population?

Question: Visualize how the daytime population changes across the Bay Area at various levels of geospatial granularity. How has this changed over time, and what are the current trends?

5. Transit Fiscal Cliff

Question: What would happen if transit service is significantly cut?

6. Builder’s Remedy

Question: How could Builder’s Remedy affect our region?

7. CalEnviroScreen Disadvantaged Communities

Question: Examine the CES 5.0 methodology and impacts and propose an alternative methodology.

8. Identifying Flood Risk Reduction from Marshes

Question: What level of flood risk reduction are wetlands providing to communities around the Bay? How are these benefits distributed?

9. El Niño and Sea Level Rise

Question: What might the impact of El Niño be on coastal flooding this year (estimated additional 1 ft of SLR)?

10. Estimating Visitation to the Shoreline

Question: How can we use free and accessible data to estimate visitation counts to the shoreline over time?

Additional Research Questions

These questions are just to assist you in choosing a relevant research topic. You can choose any other research question you like and we will help you sscope them.

Wildfire Hazard Exposure

Question: How does residential exposure to Very High Fire Hazard Severity Zones vary across demographic groups and between WUI and non-WUI areas in the Bay Area?

Data:

Note: CAL FIRE Fire Hazard Severity Zones represent long-term physical wildfire hazard, not household wildfire risk, vulnerability, or expected loss.

Affordable Housing

Question: How does the number of housing units permitted compare with the number completed across Bay Area cities? Which cities have the largest and smallest gaps, and how have those gaps changed over time?

Data: Housing and Community Development Open Data

Affordable Housing and Transit Access

Question: Does proximity to transit mean good transit access for California LIHTC developments? How does the frequency and availability of transit service vary across affordable housing projects?

Data: CTCAC data, GTFS data

Streamlined Affordable Housing

Question: Where are affordable housing projects that use state housing streamlining provisions being built, and how do their locations differ from other affordable housing projects?

Using the APR data, identify projects that used state streamlining provisions and compare their geographic distribution with other affordable housing projects. Incorporate demographic data to further analyze the effect of these bills on housing location: are there differences in where state streamlined projects are located compared with all affordable housing projects?

Data: HCD Annual Progress Report (APR) data

Measuring Access to Services

Question: How much does a neighborhood ranking change depending on how access to services is measured?

Choose any city and compare several neighborhoods using different measures, such as:

  • distance to the nearest service;
  • travel time using the street or pedestrian network (compare 15, 20, 25 minute city);
  • number of destinations reachable within a given time.

Then compare the rankings. Do the same neighborhoods come out on top under each measure?

Suggested data: OpenStreetMap or Overture Maps for networks and destinations, city open data portals for services, GTFS for transit, and Census/ACS data for an optional demographic extension.

Interested in bringing computer vision or machine learning into your project? Reach out to me and we can scope it together.

Table of contents