AI in Robotics

AI in robotics is being applied to the decisions around industrial automation: recognising a component, selecting a gripping point and adapting a task to the work arriving at a cell. MachineToolNews.ai covers the robot manufacturers, software developers and integrators bringing these technologies into factories and industrial logistics.

This page brings together our reporting on industrial robots, cobots and physical AI. Explore the applications below, then follow the latest robotics articles in the news feed.

AI in robotics on the factory floor

Vision, gripping and part handling

Part variation is a demanding test for a robot cell. A trial needs to establish whether the system can recognise the parts, choose a workable grip and recover when a pick fails. Festo describes GripperAI as software developed for bin picking. Read our Festo GripperAI report for a closer look at the application.

Machine tending and production cells

For machine tending, the whole cell deserves scrutiny. Ask how a proposed system handles loading, orientation, gripping changes and exceptions at the machine. Measure completed good parts per shift as well as individual robot cycle times. Our AI in CNC coverage follows the machining side of these connected production systems.

Physical AI and deployment

Physical AI is a term used across robotics announcements, demonstrations and production projects. When assessing a claim, identify the task, the hardware and the operating conditions. Our reporting on KUKA AMP in automotive production and FANUC and Google follows developments from suppliers working in this field.

What manufacturers should measure

Choose a repeatable task and establish its current cost before a pilot. Include programming, commissioning, tooling, operator training and support in the assessment. Test the awkward parts and interrupted cycles that occur during a normal shift. Record interventions, failed picks, changeover time and the quality of the resulting work.

For a cobot project, include the application risk assessment and the proposed operating speed in the discussion with your integrator. Ask who is responsible for maintaining the vision models and robot software, and how changes are checked before returning the cell to production.

Questions about industrial AI robotics

Where should a manufacturer start?

Start with one task whose performance can be measured. Provide the integrator with representative parts, the required throughput and a record of common exceptions. Use those conditions for acceptance testing.

How should a robot demonstration be assessed?

Ask which parts of the demonstration are available to customers, what must be configured for your factory and what evidence exists from production use. Follow the original supplier information linked in our reporting for product scope and availability.

Browse the Robotics archive for earlier stories, or explore how automation connects with AI in sheet metal and manufacturing software.

Assess force control, edge quality and production testing in our AI robotic deburring buyer’s guide.

Latest robotics and cobot news