• 03.05.14

Quantifying The Art Of Skateboarding With A Trick Algorithm

Wearable, data-generated gadgets are moving beyond runners to quantify skateboarders’ every movement.

Quantifying The Art Of Skateboarding With A Trick Algorithm
[Image: Flickr user Adriano Agulló]

Runners and fitness junkies have been able to track their progress via smartphones for years. But what about more extreme sports? Thanks to a combination of smart hardware, apps, and algorithms, the “quantified self” movement will soon branch out onto the half-pipe.


Krack is trying to quantify skateboarding the same way that running and other types of movement have been successfully mined for data. To do this, the company is adding sensors to detect movements of the board. The plans aren’t finalized, but the initial effort puts a gyroscope, accelerometer, and magnetometer into a designed case that’s plugged to the board directly under the wheels, against the truck base.

The idea seems simple enough, but Krack founder Kevin Straszburger says, “In the skateboarding world, two tricks are never the same, even 180 degrees is not exactly a 180 degrees. That means that finding patterns to develop our ‘trick recognition algorithm’ isn’t an easy thing.”

A 9-year-old beginner’s ollie is completely different from someone who’s 35 and has been skating for decades. This is a huge challenge for the team that they’ve been working on to optimize for and get just right. Ultimately, it’s separating the noise from the trick and accurately detecting movements with the board that will make or break whether the product catches on.

The initial phase of gathering data included having people perform the same trick over and over again to find the characteristics that could only be associated with that trick. The algorithm doesn’t only recognize known tricks–it also plans for the future.

Straszburger says skaters are coming up with new tricks all the time and they’re looking at ways to detect those automatically. “Skateboarders keep inventing tricks that are in fact a combination of simpler tricks. For example, when you say this person landed a ‘flip front board,’ it means they performed the sequence: rolling + flip + frontside boardslide + landing + rolling again.”

The solution is to combine the simple tricks in the algorithm to automatically adjust for the complex tricks. Fulfilling that promise, however, means that Knack needs to get enough data in the system, something the new company is working on.


Beyond the initial appeal of detecting tricks and being able to brandish proof, Knack also wants to build a community around achievements and places. This is the same type of thing other wearables like Fitbit are trying to do as well.

Tracking user movements is one thing, but then taking that data and making it relevant in a tangible way is how the newest activity tracker Moov has gained so much attention during its announcement. The Moov sensors are moving beyond raw data and providing users with features that might, at some point, replace personal trainers for a large part of the population.

For example, Moov can be attached to different sports equipment and gauge performance, even suggesting adjustments to improve outcome. Apple and Google are rumored to be interested in this space as well, but with both companies leveraging huge built-in user bases, the community and engagement level could be tremendously raised.

Krack is still in the early stages of what it’s attempting, but it’s the focus on providing users with results beyond distance traveled and top speed that makes the project ripe for the next generation of “wearables.”