Project Ideas for INFOMDSS

Explore semester-long data science project ideas involving real-world problems, live or regularly updated data, predictive or descriptive analytics, and an interactive dashboard. The ideas are intended for student teams with mixed technical experience and can be adapted to different levels of complexity. Use the examples to compare possible directions, identify an interesting problem, and develop your own project proposal.

Op zoek naar ideeën voor data science-projecten voor de INFOMDSS-cursus?

Deze pagina bevat voorbeelden van data science-projecten gerelateerd aan maatschappelijke vraagstukken, open data, machine learning en interactieve dashboards. De ideeën kunnen worden aangepast voor een universiteitsproject van een semester.

INFOMDSS Project Ideas

Before you Begin

You begin with a problem, not a solution looking for a problem. A dataset is not yet a project idea. A strong project begins with a problem, a user, and a decision that could be improved. Only then should you determine which data, KPIs, analytical methods, and visualisations are required.

Remember that these ideas are displayed as a starting point for Data Science and Society (INFOMDSS) students at the Utrecht University. These should not be considered as a descriptive project statements, and does not replace the latest version of the project guide provided in the course page on Brightspace/MS Teams channel.

Monitor Air Quality

Air-quality data is available from monitoring stations, but individual readings are difficult to interpret. This project would help users distinguish normal variation from persistent or unusual pollution events and identify locations that may require attention.

Possible users: Municipal environmental teams, schools, residents

Possible KPIs:

• Percentage of monitored hours within the acceptable range

• Current pollution level compared with the historical baseline

• Number of consecutive hours above a threshold

• Percentage of nearby stations showing the same increase

Detect Unusual Traffic Disruptions

Low traffic speed may be normal during rush hour, while a smaller decline elsewhere may indicate a serious disruption. This project would compare current road conditions with expected patterns to identify unusual delays and prioritise affected locations.

Possible users: Traffic-management teams, commuters, logistics planners

Possible KPIs:

• Percentage of journeys completed within the expected travel time

• Current speed compared with the normal speed for that road and time

• Duration of the disruption

Map Neighbourhood Heat Vulnerability

The same temperature can create different levels of concern across neighbourhoods because of differences in population, building density, and access to green space. This project would help organisations prioritise areas for outreach or further investigation during hot weather.

Possible users: Public-health teams, housing associations, community organisations

Possible KPIs:

• Percentage of vulnerable neighbourhoods covered by a heat-response plan

• Number of consecutive hot days or nights

• Percentage of older residents in the neighbourhood

• Percentage of green or shaded surface

Compare Housing Affordability

House prices alone do not show whether housing is affordable. This project would combine prices with income, housing supply, household characteristics, and location to identify areas where affordability is improving or deteriorating.

Possible users: Municipal housing teams, researchers, prospective residents

Possible KPIs:

• Ratio of average housing cost to average household income

• Rate of house-price growth

• Rate of household-income growth

• Change in the available housing stock

Prioritise Building Energy Improvements

Individual energy labels provide limited support for identifying broader renovation needs. This project would combine energy performance, building age, type, and location to highlight areas where inefficient buildings are concentrated.

Possible users: Municipal sustainability teams, housing associations, energy advisers

Possible KPIs:

• Percentage of buildings meeting the selected energy-performance level

• Percentage of old buildings without a recent energy assessment

• Rate of new or improved energy-label registrations

• Concentration of low-performing buildings within an area

More ideas to consider for your Data Science and Society project

Environment and sustainability

• Detect underperforming residential solar installations

• Compare neighbourhood energy consumption and renewable generation

• Monitor local air quality and identify unusual pollution events

• Analyse urban heat and identify vulnerable neighbourhoods

• Track waste collection, recycling, or illegal dumping patterns

Mobility and public space

• Predict public-transport disruption and passenger impact

• Identify dangerous cycling locations

• Compare parking pressure across neighbourhoods

• Monitor traffic congestion and suggest alternative travel times

• Evaluate accessibility of public services by public transport

Health and wellbeing

• Monitor environmental conditions associated with heat stress

• Analyse regional patterns in physical activity or public-health indicators

• Identify pressure on healthcare services using public aggregate data

• Compare access to green space and recreational facilities

Cities, housing and public services

• Compare housing affordability across municipalities

• Monitor neighbourhood development and population change

• Analyse accessibility of schools, healthcare, or public facilities

• Identify areas experiencing unusual nuisance or service complaints

• Compare building characteristics with energy performance

Society and economy

• Monitor regional employment and skills trends

• Compare cost-of-living pressures across regions

• Analyse tourism pressure and seasonality

• Track business activity or company formation

• Monitor progress on municipal sustainability objectives

Developing KPIs

Check out the KPI development worksheet to help thoroughly define the scope of your project.

Wat zijn goede data science projectideeën voor studenten?

Goede projectideeën combineren een duidelijk maatschappelijk of organisatorisch probleem met beschikbare data, een specifieke gebruiker en een haalbare technische oplossing.

Waar vind ik Nederlandse open data voor een data science-project?

Nederlandse datasets en API’s zijn onder meer beschikbaar via CBS, PDOK, KNMI, data.overheid.nl en gemeentelijke open-dataportalen.

Hoe kies ik een onderwerp voor een data science-project?

Vergelijk ideeën op probleemrelevantie, gebruikersbehoefte, databeschikbaarheid, technische haalbaarheid en de mogelijkheid om het resultaat te evalueren.