Jason Sherwin founded deCervo to help baseball players make better decisions. He never expected the technology to be used in life or death situations. The startup, set up by Sherwin and his friend Jordan Muraskin in 2014, was initially one of a wave of companies trying to bring brain-based training to the world of sport. Its first commercial product was uHIT, a neuroscience-driven tool for training baseball hitters.
The idea was to help players “train above the neck.” By showing them video clips and animations of different pitches coming toward them on a screen, deCervo claimed uHIT could add to their well of expertise without them actually needing to be on the field—thereby improving their ability to make the snap, high-speed decisions required when a fastball was beaming toward them at almost 100 mph.
Sport, of course, isn’t the only area where people need to make high-speed decisions under pressure. Police officers are thrown into service with barely a fraction of the thousands of hours of practice an elite athlete might have built up before they turn pro. In the United States, the average police academy training program last between 13 and 19 weeks. A high school football player in the US trains for more than 500 hours a season, Sherwin says, while a police officer may only go to the shooting range a couple of times a year. “It’s the most high-risk work possible, and we don’t provide them with the training,” he says.
While working on video training software for the NHL, Sherwin and his colleagues began to wonder whether the framework they were developing for athletes could be used to help police officers make better decisions where the stakes are a lot higher. Recently, deCervo unveiled e-Train, a product that uses the same principles as uHIT, but applies them not to fastballs and sliders but traffic stops and use of force. It is, Sherwin says, “a way of potentially solving these quick trigger decisions they can go either way, and which have much more drastic consequences than a swing and a miss at the plate.”
There are dozens of examples of police officers making the wrong decision. In 2014, 12-year-old Tamir Rice was shot dead in Cleveland, Ohio, when an officer mistook a toy gun for a real one. In 2020, 47-year-old Andre Hill was killed leaving a friend’s house in Columbus, Ohio, after a neighbor reported an SUV in the area repeatedly turning its engine on and off. He was unarmed. In April, Daunte Wright was shot dead in Minnesota when an officer allegedly fired her gun instead of a Taser during a traffic stop.
Mike Malpass, a serving police officer and former SWAT team officer based in Phoenix, Arizona, says the current approach to police training is as if a baseball player learned the rules of the sport in a classroom, hit a few slow balls in a batting cage and then was plunged straight into the major leagues. Malpass is the author of Taming the Serpent, a book about how neuroscience could improve police training, and for 15 years he’s been evangelizing about the need for officers to be trained in how to make better decisions under pressure. “Skills are great, but if you have the knowledge and you don’t know how to use it, it’s a skill you can only use on the range or in a classroom.” Training has barely changed in 30 years, he says.
“Most police training is either static—shooting at a paper target—or scenario based,” says Rudy Hall, a 20-year NYPD veteran with a doctorate in education who is consulting for deCervo. “The reality of that is limited because there’s another officer playing the role of the suspect.” Instead, e-Train uses body-cam footage from real incidents to give officers the feeling of being in a real-life encounter—but it can be slowed down and rewatched, and key facets can be highlighted so that officers know what clues to look for when put in a real situation.
In sport, that might be using the angle of an opponent’s hips to predict which way they’re going to move, says Hall, who played baseball, basketball, and American football as a child. In policing, it could be someone’s body language, or the way they angle their car when you pull them over, that acts as a hint that they may be about to flee. “It’s about diffusing the knowledge from these really subtle cues, diffusing that institutional nous to more people at once,” he says.
In a panic situation, the amygdala—a primitive part of the brain that controls our fear response— can hijack decisionmaking, bypassing the more rational parts of the brain, which can mean all the classroom training on procedures and constitutional law flies out of the window. Studies in the Netherlands found that shooting accuracy, communication, and self-defense skills all decrease when stress levels are high, and that officers were more likely to fire on suspects who had already surrendered in high-anxiety situations.
The aim of e-Train is to stimulate those circuits in a controlled environment, allowing officers to recognize and interrogate their emotions. “Body-worn cameras give us a direct data source of the stimuli that can instigate those circuits on an emotional level,” Sherwin says. “Rather than just using them for broadcast purposes for the news, we can use them to help officers identify their nervous system states, and then to make decisions in that environment.”
Other companies are developing similar products, which are being trialed by police forces around the United States. To develop its products, deCervo partnered with the NYPD, and it is now marketing its services to other police forces. Cognitive Command, founded by psychologist Jonathan Page, has had its technology adopted as part of the curriculum by a police academy in Washington state. Polis Solutions, founded by University of Washington sociologist Jonathan Wender, has trained hundreds of officers using an approach that gradually ramps up stress levels over time.
But preparing officers for stressful situations risks simply putting a sticking plaster on a much bigger problem—the tendency for police officers, particularly in America, to approach even minor interactions with the public in an almost frenzied state. They’ve been accused of brutality in their dealings with protestors, for instance, and there are regular news stories about seemingly unwarranted aggression on the part of police officers, sometimes approaching unarmed civilians shouting obscenities with their guns drawn.
US police have been accused of behaving more like soldiers dropped in hostile territory when handling arrests or protests. The militarization of the police starts with equipment—often handed to law enforcement by the military—but seeps into behavior patterns and attitudes to suspects as well. It’s also reflected in training: A 2006 report found that police academies spent 110 hours on firearms and self-defense training, and just eight hours on conflict management.
Officers sometimes rush into physical altercations, Malpass says, and then have to resort to using force because they find themselves being overpowered, when a more cautious strategy may have been able to resolve the situation peacefully. They put themselves in situations where they don’t have enough time to think.
Malpass would like to see technology being used to track police officers and spot signs of fatigue, angst and anxiety. He sees a situation in the future where an officers’ vital signs could be relayed to their dispatcher, who can advise them to take a ten-minute time out between calls if their heart rate is still too high, for instance.
Those problems can be exacerbated by racial bias. Black people are more than three times more likely to die during an interaction with the police than white people. Hall suggests that e-Train could be used to help identify when officers might have unseen bias—if they react differently to simulated encounters with Black suspects, for instance.
Some police officers now undergo implicit bias training in an attempt to make them aware of the prejudices they might hold, and how they may be subconsciously treating suspects of a particular race differently. The system of clues and shortcuts the brain uses to make high-speed decisions can sometimes let us down when it makes faulty assumptions, in the same way that an algorithm trained on bad data can spit out biased results.
But implicit bias training has been criticized for failing to tackle the problems it claims to, and there’s a risk that brain-training could fall into the same category—providing a veneer of science, while in reality doing little to tackle the underlying structural problems. By the time an officer is approaching a suspect with their gun drawn, the game has already been lost—improving their decision making at that point is important, but helping them avoid that situation in the first place is critical.
“Officers, for the most part, are starving for this information,” says Malpass. “They want anything that can help them.” That means giving them the tools to make better decisions—it might start with a brain-training app, but it could also require a sweeping change to the way police interact with people, and a reworking of the relationship between law enforcement and the public that gives them more time to think. “We’re not out to cure racism,” Hall says. “But we’re trying to keep it from coming to work.”
This story originally appeared on WIRED UK.
- 📩 The latest on tech, science, and more: Get our newsletters!
- Freedom, mayhem, and the uncertain future of Revel mopeds
- The hostile takeover of a Microsoft Flight Simulator server
- Goodbye Internet Explorer—and good riddance
- How to take a slick, professional headshot with your phone
- Online dating apps are actually kind of a disaster
- 👁️ Explore AI like never before with our new database
- 🎮 WIRED Games: Get the latest tips, reviews, and more
- ✨ Optimize your home life with our Gear team’s best picks, from robot vacuums to affordable mattresses to smart speakers