Kindred was founded with the mission of embodying AI. Without a body, AI would not be able to learn as humans do, interacting with the real world. The initial objective was to build something for a commercial use case.

Secretive Canadian Company Teaches Robots to Be More Like People

With a small bootstrapped team, we built a humanoid form factor robot that could pick 50lbs items from a 11 ft tall warehouse shelf.

Quickly pivoting after an amazingly ambitious prototype robot, we moved to a put-wall system where heterogeneous items from mixed orders could be sorted into specific batches without the use of prior item or dataset learning.

How a Human-Machine Mind-Meld Could Make Robots Smarter 

The initial concept used a more sleek concept of plexiglass encapsulating a robot arm, with an under-actuated antipodal gripper and a simplified vision system. The whole unit was able to be moved into a conference room and demoed for prospective interested customers and investors.

The concept, after finding several customers, was able to be developed into a full fledged product with target objectives. The product was able to be installed across the country, shipped turn key to customer sites and deployed quickly.

Warehouse robot startup Kindred signs up the Gap to test its first commercial product 

During my tenure at Kindred and Ocado, many additional robotic use cases which never surfaced publicly were prototyped and deployed at small scale to logistics warehouses. One of which that can be highlighted is the parcel induction cell - a dual arm, high speed system for ground sortation steps required throughout the country.
AI-Powered Kindred INDUCT Automates Induction of One Million Items with 95% Accuracy