Unlocking The Power Of Artificial Intelligence Should Be A Priority For Infrastructure Leaders Posted on : Jul 16 - 2018

Most municipal and private sector infrastructure leaders seldom if ever think about how technologies like artificial intelligence (AI) and machine learning can help improve physical systems like roads, mass transit networks or water utilities.

Although many cities are starting to recognize the importance of data, only a handful are prioritizing data collection and even fewer are feeding that data to advanced algorithms that can improve decision making. This is understandable since the public and private sector leaders responsible for our infrastructure typically have little time to think about the future while working tirelessly to ensure that people can get to work in the morning and that the water flows when people turn on the faucet.

Yet, the opportunities presented to infrastructure decision makers by the advent of data analytics, AI and machine learning are fast becoming too compelling to ignore. From predicting where repairs or new infrastructure will be needed to automating mass transit to improving project management and contractor coordination to optimizing the movement of everything from cars to commodities, digital intelligence can help cut costs, increase efficiency and enhance outcomes.

For instance, Kansas City, Missouri, in partnership with Xaqt, is using machine learning algorithms to crunch information from in-road sensors, video cameras, weather data and other sources to predict and preempt pothole development. Many cities rely on reports from drivers or have to manually review video camera imagery to identify potholes — a reactive, rather than proactive, approach that creates a backlog of repair work. Kansas City, however, is now able to predict potholes with an 85% accuracy rate, resulting in a 30% cost savings.

Pittsburgh, working with Rapid Flow Technologies, is installing artificially intelligent traffic lights that use machine vision and sensor data to adapt to road conditions in real-time rather than at set intervals. The system uses predictive algorithms to manage traffic lights dynamically and sends that data to adjacent intersections to optimize the entire signal network. This stands in stark contrast to most cities that are forced to manually program light times — a task that often necessitates time-consuming research. Pittsburgh’s smart lights on the other hand are estimated to have reduced wait times by over 40%, translating into a 25% reduction in travel time. View More