Urban renewal: How reality capture data is supporting Detroit’s revitalizationUrban renewal: How reality capture data is supporting Detroit’s revitalization

Reality capture data plays a major role in urban redevelopment decisions by providing up-to-date information that is easy to access and share. Both strategic planning and daily operations start with a complete understanding of the existing physical environment and rely on collaboration to compare and evaluate options. To help support these efforts in Detroit, the Enterprise Geographic Information Systems (EGIS) team within the Department of Innovation and Technology built Detroit Street View (DSV). The comprehensive mobile mapping program captures panoramic imagery and Light Detection and Ranging (LiDAR) for every street in the city and derives data layers such as elevation models, streetlights, utility poles and manholes—and makes the data accessible to all government departments and the public.

Imaging on a shoestring
Due to decades of urban blight, economic struggles and declining population, thousands of homes and commercial buildings have been abandoned, foreclosed and condemned throughout the city. Since 2014, Detroit has demolished more than 25,000 houses and sold or rehabbed more than 17,000. But thanks to new investment, conditions are improving.

In recent years, a citywide demolition program has replaced vacant and abandoned properties with affordable housing, green space, retail and more, while the federal infrastructure bill is helping update the stormwater management infrastructure to improve the city's flood prevention efforts. U.S. Census Bureau population estimates released in 2024 showed growth in Detroit’s population for the first time in more than six decades. Although recovery is a long process, the building blocks are being put in place for long-term success, and the DSV program is doing its part by continuously mapping the city’s 143 square miles of urban fabric.

The city’s GIS team started the innovative DSV pilot program in 2017, capturing single-frame GoPro imagery from a van to support redevelopment efforts along commercial corridors. When news spread that imagery was being collected in-house, demand for coverage of other areas to support government activities increased. A Mayor’s Innovation Grant provided funding to upgrade to a Trimble MX2 mobile mapping system in 2018, and a grant from General Motors in 2022/2023 funded the purchase of a Trimble MX50—a vehicle-mounted mobile mapping system that captures highly accurate point clouds and immersive, 360-degree imagery at up to highway speed—as well as a new vehicle to carry the system. 

“We were looking for a cost-effective method of collecting street image data quickly that could be shared and used by multiple city departments. Hiring third parties to collect data was too expensive and didn’t provide the flexibility we needed to respond to emergencies,” says Tamara Fant, Detroit’s enterprise applications manager. “Our GIS department started with one person in the basement making maps, and now we have a team of 10, including three IT technicians and six analysts.”

Detroit’s “map everything” approach focuses on capturing a lot of territory accurately enough to provide high-level information, such as the location of fire hydrants, the status of demolition work or damage from a flood or fire. To avoid the cost of separate trips or external data purchases, comprehensive data is collected in one pass and may be used to complete a blight assessment, traffic sign inventory, address investigation and topographic survey. 

“The MX50 delivers consistent positioning without ground control,” says Ted Schultz, Detroit’s Enterprise GIS team supervisor. “We aren’t producing survey grade data for engineering projects, but we are populating our GIS with current information and producing 3D data products that have many practical uses.”

Other cities have noticed Detroit’s success with the street view program. Fant and Schultz had the opportunity to share their mapping experiences and ideas in 2023 when the 21st Century Cities Initiative at Johns Hopkins University invited them to present in Baltimore, and they regularly network with public agencies and private sector entities across the nation.

User-friendly workflow
DSV provides foundational imagery and imagery-derived data layers through Mapillary, an online mapping and crowdsourcing platform. LiDAR data may be requested directly from the city’s EGIS team. All products are accessible to government agencies and the public to view and download. To ensure data is current, two full-time drivers operate the Trimble MX50 every day during suitable weather. The goal is to complete one full capture of the city every year plus multiple passes of high priority areas undergoing rapid change. Before and after shots of these residential and commercial areas are helpful to track demolition and construction progress. 

Detailed data layers showing the locations of signs, hydrants, catch basins, and other features are accessible to government agencies and the public to view and download through the Mapillary crowdsourcing platform. Image Detroit, Michigan by codgis, licensed under CC-BY-SA.

Post-processing of the MX50 data is completed in Applanix POSPac mobile mapping software, then moved into Trimble Business Center (TBC). The LiDAR point cloud is classified in TBC and exported to a GIS environment where additional refinement takes place, such as classifying and extracting features, deriving digital raster products and thinning/compressing the point clouds for distribution.

A third-party privacy algorithm is applied to blur faces and license plates throughout the imagery before it is archived or uploaded to Mapillary for public hosting. The city’s GIS team is experimenting with machine learning feature detection and asset extraction from imagery and point clouds to create additional GIS layers and expanded attribute information such as inclination angles.

Important to operationalize the data
Current geospatial information is needed to assist with better planning and asset management throughout Detroit. There are 2,000 city employees with access to GIS data on the city’s Esri ArcGIS platform and countless users accessing the DSV data directly in Mapillary or integrated applications.

The 360-degree panoramic imagery and terrestrial LiDAR captures a wide range of data related to transportation, streets, pedestrian infrastructure, urban forestry and construction projects. Multiple government agencies can access accurate georeferenced spatial data from DSV and develop custom applications by transforming it into 3D models, maps, GIS deliverables and more.

“The value of terrestrial/mobile data lies in replicating the environment from the perspective of citizens to help visualize, manage and improve their surroundings,” Schultz says. “So far we’ve captured 12 million 30-megapixel 360-degree images plus LiDAR, and we keep discovering new applications.” 

For example, the GIS team uses the classified LiDAR point cloud exported from TBC to assess road surface variation. They generate minimum ground elevation surfaces in Esri ArcGIS Pro and compare each raster cell to a neighborhood median value to identify and quantify cells above and below a normalized road surface. Digital Elevation Models (DEMs) also assist in assessing drainage and identifying road ponding issues.

The classified LiDAR point cloud is used to generate Digital Elevation Models (DEMs) that support profile creation as well as assessment of surface variation and likely ponding. Photo credit EGIS, City of Detroit.

The Detroit Water Sewerage Department (DWSD) maintains a water and sewer system with an average asset age of 95 years and operates more than 2,700 miles of water mains and 30,000 fire hydrants. DWSD implemented Trimble Cityworks, a GIS-centric asset management system that integrates with other systems to streamline workflows and provide accurate data to assist in making informed decisions. DWSD’s team references DSV products when verifying asset locations.

Smart decisions drive growth
With its automotive heritage and commitment to technological innovation, Detroit is making significant strides towards becoming a smart city, including upgrading its infrastructure with smart technologies including LED street lighting and smart water meters. A wide variety of insights and foundational geospatial data can be derived from imagery and point clouds, increasingly empowered by emerging machine learning and computer vision technologies. 

“We want DSV embedded across all government operations to leverage the cost savings and extract the full value,” Fant says. “And free access to municipal data increases transparency and digital equity for the public, which encourages them to be a part of the process.”