I love checking my Fitbit App at night, on the days I exercise. Watching the good green vibes explode off my phone when I hit the preset step & activity targets make me feel like a ninja. But when I’m sitting down and I shake my wrist a bit and the step count jumps for no reason it’s like I’ve eaten a donut and my taste buds missed the party.
A big problem
While I’m not giving up on my Fitbit just yet, this is a big problem. But it’s also exciting, because we have all the right ingredients to solve it.
People solve problems
The first ingredient, and most important, is people. People are great at finding problems and also coming up with ideas to solve them.
The second, and one we have in abundance, is data. I, and many others, have been providing Fitbit with years of data on my exercise and sleep habits. If they asked me for permission to anonymously share my data to build a better step and sleep counter I would say yes quickly.
Finding the patterns
The third ingredient is machine learning. Machine learning happens when we teach computer programs to find patterns in data. Surely within that repetitive data lurks a pattern that can tell when I’m really busting out the steps and when I’m having an animated conversation.
Dump these ingredients in a computer mixing bowl with amazingly fast processors that can find the way to improve the counter in days, not months. Then I’ll know how many steps I need to counteract those Christmas party indulgences and truly attain ninja status.
And that’s the great thing about problems.
- You’re not the only one thinking about it
- The data is out there to help solve it
- Human intelligence can guide the technology needed for the end solution
So, what’s your problem?