AUSTIN, Texas — Austin Police Chief Joseph Chacon on Tuesday outlined the results of a new report that uses machine learning to develop a patrol model for the department.
The research project was funded by the Greater Austin Crime Commission and conducted by Dr. Giovanni Circo at the University of New Haven. A large-scale survey of Austin residents was also conducted by Dr. Sean Patrick Roche at the Texas State University School of Criminal Justice and Criminology to discover community perceptions about which police services should be prioritized.
The patrol model analyzed more than 6 million officer responses to nearly 2 million calls for service, the Greater Austin Crime Commission said.
Based on that analysis, the model recommended a target response time of 6 minutes and 30 seconds for the best public safety outcomes, including the likelihood of arresting a suspect and recovering a firearm.
“I’m very excited about it – to be able to use a tool that is really evidence-based and data-driven,” said Chacon. “We pride ourselves as a police department in using data and evidence to make decisions, especially policy decisions that are going to make sense for our community.”
Based on the machine-learning model, the research team recommended a staffing level of 882 patrol positions, which is 108 above the current authorization of 774. Chief Chacon said the actual number of officers is currently “well below” the authorized 774.
Chacon said the department has seen an average of 13 to 15 officers leaving per month for various reasons, including 17 set to depart this month.
The APD's average response time is currently running in the 8-minute range, but was closer to the 6.5-minute target time in 2016, the Greater Austin Crime Commission said.
Chacon plans to meet with the Austin City Council about the report and present detailed plans to the city manager to implement patrol plans within the city budget. He said he will determine in the next several weeks if he asks the city council for any budgetary changes.
The full report can be found below.
PEOPLE ARE ALSO READING: