Sensor management SDK ‘helps robots see the world’
The company, whose stated mission is to make sensors simpler for robotics, drones, autonomous vehicles and factory automation, was founded by computer vision veterans to address what they saw as a significant software gap for sensors.
“We were working with some of the most sophisticated engineering teams in the world,” says Tangram co-founder Brandon Minor, formerly of spatial computing company Occipital, “yet they still struggled to add sensors to their platforms, and struggled even more to make them operate reliably.”
Co-founder Adam Rodnitzky, also an Occipital alumnus, adds, “We spoke to dozens of robotics companies before launching Tangram. We found that they were delaying product launches by over a year because of struggles with integrating and configuring the sensors their products used.”
Designed to address these issues, the Tangram Vision SDK includes APIs for plug-and-play integration for popular perception sensors like 3D, LiDAR, and RGB cameras, as well as IMU motion sensors. It also includes a set of tools to maximize sensor uptime during deployment by managing sensor calibration, data streams, and diagnostics.
“The flagship of the Tangram platform is calibration for multiple sensors,” says Rodnitzky. “Calibration errors are the single trickiest sensor failure mode faced by robots and AVs, and without good calibration, a robot can’t safely operate. The Tangram Vision SDK effectively eliminates those errors.”
In addition to the SDK, the company also announced the completion of a pre-seed funding round. The round was led by 2048 Ventures, with participation by Trucks VC, Dynamo Ventures, and SHAKTI.