The Grand Valley Metropolitan Council, which serves as the centralized hub of information for 39 West Michigan municipalities, has been using AI and machine learning to store, analyze and forecast data to plan for regional growth.
The council employs AI as it helps coordinate government services across the region. This includes overseeing a regional geographic information system, known as REGIS, and using transportation modeling and forecasting to let municipalities know when they should invest in infrastructure improvements.
The biggest role of REGIS is housing data and tracking it for cities, said Jason Moore, GIS Manager of REGIS for GVMC.
“We do a lot with utility lines, water and sanitary sewer — that’s probably one of our biggest roles is maintaining that infrastructure,” Moore said.
The REGIS data is stored and updated by GVMC, which makes it accessible to members in the region for city planning purposes.
Laurel Joseph, director of transportation planning for GVMC, uses a travel demand model to analyze congestion and other performance measures on the regional transportation system. Her department also analyzes pavement condition data and modeling to forecast when roads will need maintenance, as well as different investment strategies municipalities could take.
“We look at the current year, and using our modeling growth factors we try to project what the travel demand on our transportation system could look like in the future,” Joseph said.
The models can also forecast the effects on transportation infrastructure for different land uses and predict what an area could look like under development scenarios, Joseph said.
As the GVMC compiled the 2045 Metropolitan Transportation Plan, the biggest takeaway was that citizens’ main priority was improving road condition and safety in transit systems.
The GVMC is working on a way to model travel time reliability, which they have not been able to predict in the past, Joseph said. Typically, a city has a rush hour when residents and commuters can expect a longer travel time, but it becomes an issue if that rush hour period becomes unpredictable and hard to plan for, Joseph said.
“We have current and past data, but haven’t been able to use our model to forecast for the future, like how reliable will (travel time) be in 15 or 20 years,” she said.
In light of the pandemic, the GVMC’s transportation division is tracking traffic counts to see if travel will get back to pre-stay-at-home order levels. Counts show that traffic in the region is 20 percent lower than it was last year, Joseph said.
“We’re doing pre-work on a regional freight plan and thinking about supply chain management related to pandemics like this and other natural disasters to see how we can make sure our transportation system is more resilient to unanticipated changes,” Joseph said.