OLG rolls out facial recognition across Ontario for ‘self-excluder’ program
The Ontario Lottery and Gaming Corp. (OLG) is currently rolling out facial recognition software at each of the 27 gaming facilities in the province to detect “self-excluders” —people who are admitted gambling addicts.
There are five major casino resorts in the province, two of which are Casino Niagara and Fallsview (see main story), more than a dozen horse betting tracks and other facilities which offer slot machines and other casino-style games. The self-exclusion program is designed to help people with a gambling addiction by deterring them from entering one of these facilities.
Self-excluders voluntary enter the program, sit for photos for future identification purposes and sign a piece of paper saying they won’t return to casino for a period of six months, a year or indefinitely.
“If someone has an issue with their gambling and they want to stop, they step forward,” says Paul Pellizzari, director of policy and social responsibility at OLG. “They can self-exclude in a couple of different ways now. Most people do it at a gaming site while they’re gambling. They talk to a security person, who registers them in the program and does all the administrative stuff.”
Unfortunately, the majority of people who enter the program are likely to relapse. Pellizzari estimates that 50-70 per cent of self-excluders will try to enter a gaming facility before their term of self-imposed exile is over. There are approximately 15,000 enrolled in the program, giving the OLG the task of spotting thousands of people who aren’t supposed to be in their casinos.
The old method was to get casino guards familiar with the faces of self-excluders who had gambled in their facilities in the past. The new method is to check the face every person who enters an Ontario gaming establishment against a database of known self-excluders.
Using cameras posted at casino entrances, faces are scanned using iGWatch Facial Recognition Software, part of the iTrak platform from Oakville, Ont.-based iView Systems. If a match is found, the data is encrypted and security is alerted that a potential self-excluder is on site. The rest of the data — images of patrons that don’t match the database — is automatically deleted. The system had to satisfy the Ontario Privacy Commissioner Ann Cavoukian, who is concerned about how Ontarians’ private data is used or stored. “We wanted her blessing on the privacy protection features of the system.”
If a person is self-excluded from one site, they’re excluded from all 27. The facial recognition database is stored centrally, but each facility has two servers on site to manage data locally. The cameras used to detect faces are optimized specifically for that purpose.
The system was tested vigorously before it was deployed in order to confirm the system’s accuracy and actual rollout began earlier this year. Pellizzari estimates that all 27 sites should be up and running by early June.
It’s too early to ascertain a success rate, but “in these early weeks, our detection rates have gone up dramatically,” says Pellizzari.
If detected, self-excluders are escorted from the premises and time is added on to their self-exclusion term. Most go quietly, embarrassed or upset that they have given in to their addiction, says Pellizzari. In some cases, they may face a trespassing fine, but the goal, he says, isn’t to punish people but to help them.