If you want to fire off an instant debate in technology, start a conversation about privacy. Everyone has an opinion, and it’s an incredibly complex issue. Setting aside the privacy debate for a moment, I want to briefly put a spotlight on the benefits.
And just so we’re on the same page . . . in this post, facial/body recognition refers to the ability of a computer to understand what it’s looking at through a video lens. In other words, a computer can recognize that a person is sitting, jumping, and that it’s Tom instead of Bob.
SOLUTION #1: Predicting behaviors before they happen, rather than reacting
I can’t tell you the number of calls I’ve received from school about Calvin, specifically spontaneously going after another student, or suddenly running out of the bathroom.
The reality is that like most every other child (and adult), he had body and face movements that probably would have telegraphed that he was escalating. As parents and caretakers, we learn and know the unique signals for each of our kids. In the same way, machine learning models can learn those same patterns and potentially alert staff, or even intervene if possible.
SOLUTION #2: Providing care at scale instead of solely relying on a caretaker/patient ratio
One of the most powerful potentials for recognition technologies is the ability to provide another pair of eyes to an already-overworked staff. According the AARP Caregivers and Technology study, in 2020, there are expected to be 117 million Americans who need care. They are also forecasting 45 million unpaid caregivers and 5 million paid caregivers. The reality is that we have no choice but to look to technology to fill the gap.
SOLUTION #3: Tackling the unique challenges of each person instead of a one-size-fits-all approach
When I visit Calvin’s classroom at school, it’s staggering to see the disparity between all of the kids. Each one is completely different, with completely different abilities and sensory issues. If the computer can recognize a potential behavior in a child, they obviously recognize who that child is and they can provide unique support for that child. One child may need to go sit on the beanbag, while another may need to listen to music. If the system knows where and who the child is, the system can distinctly address each child where they’re at.
There are some really cool technologies within AWS that provide great promise in this area. This is definitely one of the areas I’ll be personally tackling over the next couple years.