Police interactions are complex. They aren’t always clear or easy to dissect, even after the fact and with the benefit of footage. But one group of researchers believe they can use machine learning to dissect body cam footage and help police departments train their officers.

The spread of body-worn cameras (BWC), now worn by an estimated 20 percent of all police officers in the US, has enabled the collection of a trove of video footage of police interactions with people of all ages, races, and ethnicities.

The data is already present. The problem is how to turn all of that data into actionable information. Researchers from Washington State University (SU) partnered with a maker of police body cameras (Axon) and are now using artificial intelligence (AI) and machine learning (ML) techniques to gain insight into police behavior.

The study, being conducted by David Makin and Dale Willits, both professors at WSU, uses data analytics and algorithms “to reduce risk, enhance officer performance, and improve officer health and safety by converting body-worn camera footage into actionable data.”

Makin and Willits’ studies coded over 100 variables that provides insight not only into the occurrence of the use of force but also why, how, and for how long force was applied by police on various groups in relation to their level of training.

Information gleaned from these studies will help police departments help their officers train their officers better.

In an interview with PoliceOne, Makin said:

“We assume all our training works, but if it doesn’t change behavior in the field, then why are we doing it? If it is “check-the-box” training, we are wasting money and time and police officers will view training as just perfunctory. At the end of the day, training has to impart new skills and, with this research, we can validate if training is working.”

In policing, you can do everything by the book and still have a poor outcome. You can do everything badly and somehow have a good outcome. If we can study or analyze what takes place in interactions and then develop training around it, we can use that as part of risk management.

We forget that police interactions can be inherently complex. Many departments wait for something bad to happen and then they go back and hope they have footage of the incident to figure out what transpired. The CSI lab facilitates the study of policing beyond merely the outcome to understanding what took place in an interaction. We are trying to give agencies the technology they need so they can use this footage to identify if officer training is working, when training needs to occur and when officers need refresher training, all while improving the overall state of police community interaction and officer health and safety.

Makin also added that the automation of AI-powered video reviews will also help police departments identify potential problem officers since the algorithm will automatically flag videos that need further scrutiny.