stand cover

Matta

Stand: P198
Digitalisation & Automation
Test & Quality Control
Innovation & Sustainability Trail
Inspection & Measurement
https://www.matta.ai/

About us

Matta is creating industrial AI for factories. We are a group of engineers, scientists, and builders developing manufacturing foundation models (MFMs) and a general-purpose manufacturing OS to improve the sustainability of manufacturing today and to unlock the unimaginable technologies of tomorrow.

We have two core product offerings which are accessible through Matta OS:
Blue - our generalist computer vision AI for automatic defect detection, set up in 24 hours, to reduce waste and improve quality from the first day of installation.
Grey - our causality-driven AI for closed-loop control to automate machinery throughout the factory to improve yield and unlock new design possibilities.

Matta was founded at the University of Cambridge by a team of AI and manufacturing engineers with experience from MIT, Imperial, Google X, BBC R&D, and Microsoft. We are hiring across engineering, AI research, and business development!

Products

product-image

Grey - AI for QC, automation, and process control

Grey-1 is our next-gen AI brain for process control and closed-loop feedback. Capable of advanced error detection, manufacturing parameter prediction and optimisation capabilities within 24 hours of setup, Grey continues to learn from every part it makes. This aids manufacturing of high-quality goods while minimising energy use and environmental impacts. Grey also empowers the workforce, equipping it with the necessary skills to not only fill current skills gaps but to excel in a future where adaptive manufacturing and sustainable practices are widespread. A fundamental challenge with AI systems is that they typically require extensive examples of the things they are supposed to do before they perform effectively. This can be challenging in manufacturing where data can be difficult or resource-intensive to gather and store. For instance, some defects might be critical but not occur often. Or it might require highly-skilled engineers to annotate images with the things an AI system needs to learn, which is unfeasible given the amount of examples an AI system will typically need to see. Grey’s core and patent-pending technology is capable of automatic data labelling, minimising the need for laborious and slow data annotation or replication of rare failure modes to train models. Grey can also learn the causes of defects, rather than simply identifying correlations. This allows it to build a better understanding of the underlying physical processes, reducing the amount of data needed for the system to work and making it robust to changes in the environment, process inputs or product mix.

News