Frontier
Skild AI demonstrated a self-play training approach that lets robots develop physical skills without human demonstrations, using its S1 robotics foundation model to learn soccer over more than 140 years of simulated play. The company says it could extend to factory and construction skills.
A research demonstration, not a deployment; worth watching, not yet worth budgeting.
Agents
Automation World argues a useful factory AI agent does not approve automation; it helps engineers discover what could go wrong before a machine is installed.
Use agents to surface failure modes in commissioning, and keep the go/no-go with your engineers.
Workforce
RTInsights argues that a real-time model trusted enough to act without human review is important enough to fund the monitoring that proves it still works.
If a model acts without review, budget its monitoring as part of the deployment, not an afterthought.
Workforce
Schneider Electric's Neil Smith and TwinThread's Andrew Waycott answer audience questions on AI ROI, shopfloor training, knowledge transfer and accountability. They cite workforce resistance, named as a barrier to AI scaling by 25.7% of CPG companies.
Settle who is accountable when the AI acts before you scale it; resistance shows up as a stated barrier.
ForgeShift Insights goes out weekly across three editions.
Subscribe on the insights pageWe use only essential cookies. We don't use advertising or tracking cookies. See our Privacy Policy.
