Rand McNally has rolled out a new Practical AI framework within its SafetyDirect platform it says uses machine learning to measure driver and vehicle performance data.

It says the new offering gives it the ability to better direct fleet managers and safety directors on where to focus their attention.

The first feature to be rolled out is Automated Event Ranking (AER), which automatically classifies and prioritizes safety events. Rand McNally says in testing the feature produced a 70% reduction in the number of safety events requiring a manual review.

“Data is only as valuable as it is actionable,” said Andre Tokman, global head of data science for Rand McNally, in a news release. “The idea is to take big data and make it feel smaller and more useful — not just overwhelm our customers with more of it.”

The company says AER ranks safety events such as sudden braking, lane departures and other behaviors, minimizing false positives and allowing managers to focus on the most urgent issues.