Detecting potholes for better citizen experiences

Using machine learning to improve road maintenance

Reducing the 32,000 man-hours through data-driven strategies for the public good

The safety of residents and city beauty is a key focus for state & Local governments to attract tourists and maintain resident satisfaction. The City of Memphis collaborated with Google Cloud and SpringML to confirm this approach to make this a priority to improve their cities’ road pothole condition and vacant properties. With an aim to keep the streets well maintained, reduce concerns over the number of blighted properties, and create an improved experience for citizens and visitors.

City of Memphis Team

Our goal is to become a smart city, and technologies such as Google Cloud Platform and SpringML put us ahead of the game. Google understands data, and there isn’t a better company to help us analyze our data resources for actionable insights.

Jim Strickland

Saving the city $20,000 annually by using cutting-edge technologies

The City of Memphis has over 6,800 lane-miles of city streets and 15,000 vacant properties owned by out-of-town investors creating a direct impact on efficiency and perception of residents that the government was doing its job efficiently. The dependence on citizen’s reports, via the 311 app, only covered about 20 percent of the potholes and blighted properties, leaving many more unreported.

The City of Memphis Mayor, Jim Strickland, and CIO Mike Rodriguez were required to find a technological solution to overcome these toughest public works and urban planning problems. Google Cloud recommended conducting a machine learning proof-of-concept (POC) with SpringML.

Memphis is focused on easy living, and we want to do everything we can to keep our citizens happy. Working with Google and SpringML to reduce potholes and urban blight using machine learning and artificial intelligence was an easy decision.

Mike Rodriguez
CIO, City of Memphis

Enabling Memphis to detect potholes and vacant properties with over 90% accuracy

The pilot analysis for City of Memphis potholes kicked off by training TensorFlow models for ML object detection and differentiation of potholes, manholes and other objects by using pre-configured Google Cloud’s AI Platform Deep Learning VM Images on Compute Engine. SpringML coordinated, tested and enabled cameras on public transportation vehicles, and developed a user interface to collect the visonal data of potholes images to be ingested into their 311 ticketing process automatically. For the pilot and to train the ML model, these cameras were mounted on city buses and code enforcement vehicles to minimize costs. The model for the potholes was trained to classify them by width and depth, and share the information with the street maintenance crew via the 311 ticketing process.

The solution also included imported routes, potholes, and paving data along with geolocation data from ArcGIS and Google Maps into BigQuery for a better understanding of street conditions and the proximity of potholes to one another.

For vacant properties, the implementation of the BigQuery model helped in analyzing city property records, tax records, 311 reports, and third-party survey data on-demand to predict where homes are starting to become run down and where neighborhood decay is most likely to occur. The SpringML team created a pilot analysis with Google Cloud Services to begin vacant property protections and developed a user interface tool to interact with the model’s results. The model was trained to proactively and comprehensively identify and manage blighted and substandard properties.

Building a better future for Memphis’s 652,000 residents

Our collaboration helped the City of Memphis provide its residents with well-maintained roads, city appearance, and increased efficiency of Public Works road crews. Additionally, the video analysis gives further visibility (add ons) into issues that weren’t previously noticed, such as curbs, gutters, and manhole covers that had been mistakenly paved over and needed to be excavated.

Now, with this new technology, Memphis will be able to make a significant difference in the efforts to proactively and comprehensively identify and manage blighted and substandard properties.”

Code Enforcement with better data-driven detection mechanisms enables the city to also identify cases where homeowners are not physically or financially able to keep up with the challenges of homeownership and make them aware of resources that are available to assist them. Memphis Code Enforcement can do a better job of finding people living in derelict properties that pose hazards to inhabitants’ health and safety, and help them fix those problems or find a new place to live. The city of Memphis has seen a 97.5 percent accuracy rate in analyzing predicted trends to address high rates of abandoned and blighted homes.

SpringML and Google Cloud’s expertise is proving the viability of a cost-effective, cloud-based machine learning model that the city can look into for further improvements to city services to build a better future for its 652,000 residents.