Faster Than Mobileye: New Sensor Company Uses Lights For Positional Tracking
What it Does
In industrial Talpiot in Jerusalem, a tech company called SixdofSpace has figured out a way to achieve “ultra-fast optical tracking” by using existing lights in any setting as “positional beacons.” Without using a camera or external sensors, this tech advances far past other positional tracking systems. Mark Goldfarb, CEO of the company, spoke with me about the trajectory of the company.
Focusing on Practical Uses
The new technology is a game changer for many industries, but right now, Mark explains, the company is focusing on three specifically: AR/VR (Augmented and Virtual Reality), Automobiles, and Drones.
Augmented and Virtual Reality
Traditional AR/VR headset manufacturers generally have two options: one, to use external sensors to track head movement, and two, to use inside-out, image-based tracking systems. Needing to use external sensors is limiting and remains a significant barrier to mainstream adoption of AR/VR headsets. Having to install sensors in specific rooms means that users cannot use headsets elsewhere. The alternative method, inside-out image processing, is quite slow, requiring a computer to run complicated algorithms on millions of pixels. Headsets that use this method end up with a significant lag in their response to head movement.
SixdofSpace’s new sensors allow for AR/VR headset companies to choose a third option, one that enables headsets to fully track the six degrees of freedom of a user’s movements with only a 1ms delay in response. Mark told me that he once pitched this application of his technology to an executive behind HTC’s system and that he was impressed with the idea. After all, who would want to install sensors in a room to use a headset?
This particular application of optical tracking tech has a clear market waiting. When Facebook’s Oculus Go was released, there was significant criticism that it “doesn’t support a true six degrees of freedom.”
Automobiles
The sensors also have high potential to be used for collision avoidance. Since they quickly identify approaching lights, they could easily be used to monitor surrounding headlights and taillights. The “optical compression technique” that SixdofSpace uses is able to capture multi-megapixel images at an astonishing frame rate of up to 1000 frames per second. For comparison, self-driving tech giant Mobileye’s real-time image processing EyeQ2 vision system processes images at 15 frames per second.
Drones for Inventory
Large companies with manufacturing plants and warehouses could also benefit from SixdofSpace’s sensors. By using the light sensors, companies can automatically keep track of staff movement in hazardous areas and robotic systems in use.
Other
Of course, Mark told me, “we’ve been speaking to the army,” and simply left it at that.
Community Interaction
The SixdofSpace team interacts heavily with the Jerusalem startup ecosystem. It has students from Azrieli College of Engineering work on their final projects in their lab, attends MadeinJLM events, and occasionally connects with business networks via Hebrew University. The entire team also participated in ALYN Hospital’s “PELEthon” and produced a sensor for PELE that can assist in diagnosing swallowing problems.
Watch the company’s promotional video and check out their website for more information.
