As a company that creates custom made artificial intelligence video analytics solutions for our clients, we've had some pretty unusual requests over the time. We've come to understand that each client has particular problems no one else does and that every industry has certain idiosyncrasies you and I cannot comprehend unless we work there.
So, we've come to expect the unexpected when we sit down with a client and ask them: What are the problems they are facing that we can solve with video analytics?
But a while back, one of our clients [a big medical device manufacturing company] replied to that questions with this… women's high heels.We immediately started to laugh as we thought a joke was imminent, but he was dead serious.
Understanding the problem
What we didn't know at that time is that there are around 120,000 foot and ankle related injuries reported a year according to the U.S. Bureau of Labor Statistics.Up to 75% of these injuries are reported because of the use of high heels or improper footwear.
For each injury the company has to spend thousands of dollars in medical treatment and follow up, workman's compensation claims and increase in policy premiums, legal claims and costs, response to incident time and costs, reporting of incident time and costs, productivity interruption, potential damages to equipment and materials and more.
An injured employee will take an average of 6 days to recover from the injury. The Occupational Safety and Health Administration (OSHA) can also fine the employer a minimum of $2,300 to as much as $70,000 per incident according to section 29 CFR 1910.136 on occupational foot protection.It all adds up to a massive, endemic and a very expensive problem all industries have.
Fixing the problem
To try to curtail this, many companies have instituted policies were high heels are prohibited or a maximum height of 4-inch shoes are allowed.
The problem is that this is extremely difficult to enforce those policies especially in places where you have a high volume of employees. Assigning personnel to watch the footwear of every employee that comes into the site is not practical or reliable and is very expensive.
Thanks to our client representative, a Safety and Risk Management Officer, we now were aware of the magnitude of the problem and the benefits of creating an automated video analytic solution for it.
An artificial intelligence solution
Once we understood the scope and complexity of the problem, we had to come up with a solution. Training our artificial intelligence (AI) platform to recognize shoes over 4 inches tall would not be an easy task.
You see, we train our video analytic platform's AI the same way you were trained as a baby. Your parents or teacher showed you a picture or a dog; told you it was a dog; and you slowly earned it was a dog by repetition.
This is the same principle on how we would train an AI. For it, we would need thousands of pictures of shoes. Initially, we hatched a plan to raid shoes from our wives closets to take pictures. Thankfully better heads prevailed as we realized there are thousands of shoe images on the web we could borrow.
So, after a great deal of arduous work making datasheets, and development of specific video content algorisms, we were able to train our video analytic platform to identifies and classify shoes that are more than 4 inches high and alert the employer in real time.
Our Optical Pattern Recognition, as decreased work-related foot and ankle injuries by 90%. The small investment the client made for this development is paying huge dividends by saving them time and potentially millions of dollars eventually.
All because a simple conversation about women's shoes. So, if you have a problem like this that you think doesn't have a solution, you may be surprise how much ai video analytics might help.
Guardian Systems is an artificial intelligence video analytics solutions provider for businesses and homes. Our video analytics products, software, and solutions can be integrated to improve areas related to security, operations, human resource management, business intelligence, data management, safety, risk management, and more.