Home
Check the Project Board to see overall development status.
Check the Milestones to track progress on major pieces of work.
A healthy building has a complex ecosystem of functional parts—an exterior shell, a foundation, plumbing, electrical wiring, gas lines, temperature control, ventilation, and many other features for the health, safety and comfort of its occupants. Keeping a building healthy requires constant monitoring and maintenance, as its materials break down with use and age. It can be surprising how fast a building can develop major problems that are not apparent from the outside. At worst, buildings conditions can cause serious injury or even death to people who enter.
Local governments have a strong interest in encouraging owners to maintain their buildings. Knowing the health status of a building is important for the function of many departments, especially those like Police and Fire who enter buildings to protect communities from building-related problems. However, governments do not have the resources to proactively visit all buildings to determine their status.
As a result, government officials have a need to quickly compile information about a building’s health from external sources. Public datasets exist on crimes, code violations, vacant buildings, legal proceedings, citizen complaints, delinquent property taxes, and building permits. However, there is no single source to aggregate the data and provide an easy way to visualize or interpret it.
A person interested in getting an overall picture of a building’s health must go to a plethora of different websites. A typical search could involve visits to over 10 sites, each requiring a different search method. Learning the process takes time. Making sense of the data requires expertise. And inside city government, employees must get access to and master archaic data sources to see more detailed building data. The process is resource-intensive and cumbersome, which results in a few experts dedicating much of their time to handling the most pressing building-related issues within departments.
There is a need for an interactive mapping tool that conveys information about the health of buildings in a simple way. It would make a big impact in the day-to-day work of public safety departments. For example, the Police Department could use such a tool when investigating and protecting a community from crimes that tend to occur in and around buildings with certain predictable characteristics. The Law Department could use it when determining strategies for building safety litigation, allowing for more informed prosecutorial decisions. The Buildings Department could use the tool to expand and enhance its problem landlord list, which would allow it to better target building owners whose neglect results in dangerous living conditions for Chicago residents. Also, the Fire Department could use the tool to make decisions involving firefighter safety when entering dangerous buildings.
Consolidating building data requires a way to link records that are associated with a building. Building identification is a complex task, as property is often referred to in different ways, such as address, geographical coordinates, or tax ID number. Further complicating the task is that buildings can face two or more streets and have long address ranges. There is no established protocol that links buildings across these identifiers.
Fortunately, the City and County have building and tax parcel footprints on their public data portals. Also, all city datasets with building data are geocoded with latitude and longitude coordinates and can be plotted against the footprints. However, to link all records to the correct building, this project must include the development of a way to handle coordinates that fall outside the footprints.
The goal of this project is to create a simple visual interface to explore the health of buildings in Chicago. This will be done by classifying buildings using “health” metrics, similar to vital signs, which will be selected using a combination of buildings expertise and data science methods. The classification will be displayed visually on a map in OpenGrid (and internally in WindyGrid) and will derive from underlying building records. The data will also be provided in table format so that a user can see what data supports the classification. The raw building data that drive the classifications will be consolidated from various sources into a single data warehouse, and will be linked across datasets using building footprints and a machine learning model for linking coordinates with the correct buildings.
- Police department
- Make sure that school Safe Passage Routes are clear of dangerous buildings
- Aid a special unit of police officers who target buildings that may harbor criminal activity
- Law department
- Determine which cases to dedicate prosecution resources
- Build evidentiary support when asking judges to impose increased penalties against repeat offenders
- Building department
- Determine which buildings are being neglected and are unsafe
- Identify who owns the most troubled buildings
- Fire department
- Determine which buildings are likely to have dangerous conditions
- Use data when deciding if a firefighter should take extra precautions when entering a building