Industrial & Robotics AI

Software and ownership for physical AI’s supply chain

AIC acquires precision component manufacturers and applies its own decision software to improve factory throughput, targeting the industrial supply chains behind physical AI.

Flexible factory robots without programming each new part

Cohesive combines AI vision, sensors, and robot arms to automate sanding, polishing, and welding across variable parts without repeatedly programming new toolpaths.

AI controls the factory as new materials scale

CubeFabs builds modular semiconductor factories with AI inspection and real-time process control, helping next-generation materials move from laboratory experiments to reliable production.

One intelligence layer for mixed warehouse fleets

Destro coordinates robots, people, and warehouse systems through a shared AI platform, pairing fleet orchestration with manipulation models trained from human demonstrations.

AI joins supply chain intelligence with procurement action

Exiger applies AI to supplier, product, and risk data, helping organizations map supply chains, automate procurement, and manage compliance and resilience.

Build the robotic workforce for orbital industry

Icarus develops dexterous robots for space operations, using initial teleoperation and manipulation data to build toward autonomous cargo handling, maintenance, and manufacturing.

AI connects and runs consumer goods production

Keychain connects consumer brands with manufacturers through AI, then extends into production planning, inventory, purchasing, and traceability for packaged goods operations.

Dexterous hardware designed around AI control

Kyber Labs develops robotic hands and artificial-muscle actuators designed for AI control, enabling low-cost dexterous manipulation and embodied learning data collection.

Robots and craftsmen make elaborate stonework scalable

Monumental Labs pairs AI-enabled robotic stone fabrication with human carving, helping artists and architects produce sculpture, ornament, and structural elements at larger scale.

AI that understands the built world's project data

Nomic builds domain-specific AI agents that understand drawings, specifications, and project records, automating review and coordination for architecture, engineering, and construction firms.

Simulate complex systems to accelerate scientific discovery

Pasteur Labs develops simulation intelligence combining machine learning, scientific modeling, and accelerated computing to explore physical systems and test technologies before deployment.

AI directs the kitchen's next batch

PreciTaste uses AI to forecast restaurant demand and direct kitchen preparation, helping teams decide what to cook, when, and in what quantity.

Materials discovery in a self-driving laboratory

Radical AI combines materials models with robotic laboratories that design, synthesize, and test inorganic materials, feeding experimental results back into subsequent discovery cycles.

Humanoids that learn from real industrial work

Reflex builds wheeled humanoid robots for warehouse and factory tasks, combining robust hardware, human supervision, and AI policies that improve through deployed work.

Industrial AI that customers can own and operate

Structify builds AI applications for manufacturers, distributors, and builders, connecting operational data on a secure foundation that customers can run and extend themselves.

Teach robot hands the fine motor skills industry needs

Tangent Robotics combines dexterous hands, tactile sensors, and learned control policies to automate precise manipulation tasks such as assembly, wiring, and connector mating.

An AI operating layer for industrial work

Traba combines industrial staffing with AI agents that automate supply chain workflows, analyze operational data, and support decisions for warehouses, manufacturers, and logistics providers.

Practical robots for repetitive warehouse work

Ultra builds general-purpose AI robots for repetitive industrial manipulation, helping warehouses automate sorting, packing, and material handling without complex integration projects.