This Quantified Skateboarding Gadget Maps Your Moves In 3-D

A new startup is the latest to bring the perks of the quantified activity movement to skateboarding.

Skateboarding looks so much cooler in video games–you get to see everything from all those crazy camera angles. But what if you could bring that perspective to real life? One startup thinks you could by adding a piece of quantifying hardware to the bottom of skateboards, giving skaters a more complete look at the tricks they’re attempting and mastering.


Syrmo, a HAXLR8R-backed startup, is launching its Kickstarter campaign today to bring motion-tracking hardware to skateboards for an initial price of $99. It will also allow skaters to not only view their own action as a 3-D animation, but take video that automatically tags each trick and selectively adds slow-motion effects. You know, just like in the video games, but starring you.

“The idea came while we were playing Tony Hawk’s Pro Skater and we realized it would be awesome to have a device to make the video game real life,” says cofounder Matias Fineschi.

Attaching special riser pads–the piece of plastic between the wheels and wood board–that contains sensors allows the capture of skating data. The hardware includes a gyroscope and accelerometer which can also track airtime, height, distance, and pop force.

The hardware and app combo can also capture and tag video. Algorithms detect tricks being performed and automatically convert them to slow-motion. This automated process eliminates the need to fiddle with manually filming the action and adding effects.

Quantifying The Skateboarding Experience

The 3-D renders displayed in the app aren’t pre-set models, but actually display the motion each individual board takes. Each trick also takes less than a second to render its own 3-D model.

“We created a 3-D model of a skateboard with a CAD tool and then exported the associated STL file. Currently a webGL engine loads the STL model and then a physics engine models all the board movements,” Matias explains.


“As iOS (and Apple) is not longer offering support for webGL in future devices we are changing the webGL engine into our own native libraries which are coded in C++,” he says. “The software team will begin to develop this libraries in about two months. This libraries can be compiled either for iOS and Android since it’s native code C++.”

As far as how those tricks are initially detected, that’s the complicated part.

“We started with a database of around 2,000 tricks, which isn’t much, and those included the three canonical tricks–each of them is compose by rotations on each of the 3-dimensional axis–the ollie, kickflip, and shove-it,” says Fineschi. “Most other tricks are made by a combination of these tricks, though not all. That’s why we focused first on those three tricks to see how accurate our algorithms could be.”

“With this data gathered by our sensors we were able to develop an algorithm which could identify a trick with 98.5% accuracy, which is just the start,” Fineschi says

Improving Skate Videos

In addition to displaying a 3-D display of the skateboard in motion, Srymo is also trying to improve on the video experience.

“The device attached to the board runs an algorithm which detects when the skater performed a trick, and then sends the data via Bluetooth to the phone. The app will then apply the algorithms to see if the data received was effectively a landed trick or not,” Matias says.


“If a trick is performed the algorithm allows the app to trigger slo-motion on the exact moment you perform a trick,” he explains. “The app will also crop the video to avoid heavy data on your phone and make them easy to share. That’s how we can create a cool video in three secs, with only a few taps.”

The team of three software engineers and a mathematician have been working on machine learning trick algorithm for the last six month. It’s not enough to come up with a defined set of rules, because skateboarding is constantly changing and tricks are always evolving.

“You can never throw a dart to the exact same place,” Matias says, describing the data output of the same trick done by different people.

The difficulty of detecting tricks isn’t unique to Syrmo, however. The startup Krack ran into the same problems while developing a similar device to quantify skateboarding. The conclusion both companies have settled on is that no two tricks are the same. Ultimately it becomes all about cutting out as much data noise as possible and then determining which trick the real-time data matches the best.

For runners and gym buffs, there have long been an array of activity tracking apps available on the market. The likes of Krack and Syrmo are opening this universe up to skateboarding enthusiasts–and with both products still in development, they’re only getting started.