Future of Police and Security Sector pt. 2

Work of Tomorrow

Tom Joyce is VP for Business Development at Vigilant Solutions, an image analytics company serving the security industry. He draws from past experience working for the New York City Police Department, from 1986 to 2006. He first started working at the beginning of the crack epidemic when crime rates were on the rise. The department shifted from being more “laidback” and “reactionary” in the late 80s through the early 90s, to more proactively policing with a greater sense of accountability in the rest of the 90s until 2001, when the focus shifted to counterterrorism after 9/11. According to Mr. Joyce, Vigilant Solutions focuses on three domains: It supports its clients by creating “actionable intelligence” by using pictures of license plates, facial recognition technology, and ballistics analysis. In the case of facial recognition, Mr. Joyce clarifies that there is “no absolute determination” that a person in an image provided by public security, whether a witness or suspected offender, is in fact the person in a previously acquired image. Instead, facial recognition algorithms produce a list of candidates to the end user, the investigator. The investigator in turn looks for unique markings like scars or tattoos to confirm that the person in the queried image may be the person from the list of candidates, and then, the investigator checks if that person had been available to commit the crime (e.g. whether or not they were incarcerated at the time), and whether or not their method of operation was consistent with the crime alleged. At that point, after approval by peer review by other investigators, the lead is provided to the main investigator. A probable cause for arrest still needs to be established, independent of facial recognition. Vigilant Solutions’ business model for facial recognition has two parts. The first is an “online cloud-based environment” for investigators to upload probed images of unidentified people and generate a list of potential candidates. The second involves taking parts of streaming video and running it against a gallery of images of high-threat people; investigators will receive an alert when there is a potential match. Mr. Joyce clarifies that his company’s technology is not a tracking technology, as we often see on TV shows like CSI. Instead, the technology provides data points for investigators to follow up on. For example, a license plate reader takes two images, one of the license plate and one of the vehicle, and collects basic data like date, time, and location. Immediate analysis of the images provides further information for law enforcement, such as whether or not the vehicle is on the National Crime Information Center hotlist circulated by the FBI. This list includes over 300,000 vehicles, for not only criminal threats but also Amber and silver alerts. Mr. Joyce comments that his company is growing and continually looking for new opportunities, keeping in mind the community’s concerns about privacy. There are some areas in which his company is not involved (e.g. enforcement; speeding cameras), but its technology does allow for integration, as in the case with connectivity with other license plate reader cameras. Looking ahead, he sees the potential to decrease the time taken for an investigation, especially with dynamic video analytics, which translate hundreds of hours of footage to specific moments for analysis.