Ambi Robotics: US pick-and-pack specialist receives further funding

There is another major funding to report: Ambi Robotics is a "special" provider. The startup around a former professor at the University of Berkeley combines self-developed optics with its own suction-based gripper, container and AI operating system. The customer thus has to decide for "all or nothing".

How it works: A depth-sensing camera first looks into a bin of items and analyzes the items. After determining how best to grip the item, the robot picks up the item with its suction pad, holds it up to a barcode scanner and places the item in a bin. The system then alerts a human operator when a bin is full and ready to be packed. Reliability is said to be 95%, which currently strikes me as still borderline. A USP could lie in the particularly simple learning of new items. This is important, for example, in the case of changing offers, but also in the case of initial acquisition: anyone who has to teach in 10,000 different parts, for example, will appreciate it if only 5 or 10 images have to be taken per part instead of 100.

As a result, the CEO says, "Our secret sauce is our sim-to-real technology, where we do all the machine learning within the simulation - that's what really sets us apart"(source).

Following initial funding of $6.1 million at the end of March 2021, $26 million has now been raised.

A little cross-selling: By the way, the startup VisCheck, which I support, manages - also by means of simulation (but certainly a simpler one) - to get by with about 10 shots/part in pattern recognition. If you are interested, please call or email me.

Are we networking? LinkedIn
-> To the Cobot group on LinkedIn(Link)

On my own account/ Advertising
The author of this blog is significantly involved in the AI/ robotics project Boost-Bot. He advises robotics companies and investors on market (entry)/business development and funding/subsidies. The standard book (though already 2 years old) about cobots is also written by Guido Bruch. More about him can be found here.