Ramblr Inspect and BASF eliminate flange failures before they cause downtime

In chemical production, a single flange installed incorrectly can cause leakage, trigger a production halt, or create a safety incident. Failure points are hard to catch before they become costly, and when they do occur, the knowledge needed to prevent recurrence rarely makes it back into training.

• Improper flange installation causes leakage, downtime, and safety risks

• Manual QC processes unable to scale across thousands of components

• Failure detection dependent on individual operator experience and attention

• Operational knowledge from frontline workers not systematically captured or applied

• No structured way to identify recurring error patterns across the assembly process

Domain-specific AI models, trained on videos of BASF's own assembly process, identify common error cases and failure points across thousands of flanges. Potential error-prone installations are automatically flagged for targeted inspection – so QC teams focus effort where it matters most. Teams access relevant video evidence in one click and use natural language to locate key moments – without scrubbing through hours of footage. And because the system learns continuously from what frontline operators encounter, every inspection improves the next one.

• AI models trained on BASF's flange assembly process – detects deviations and common error cases

• Automatic flagging of potential error-prone installations for targeted QC review

• Video-based failure detection across thousands of components at scale

• Natural language video search – locate key moments without manual review

• Guided inspection workflow supporting frontline operators on the ground

• Continuous learning from operator experience to sharpen detection and improve training programs

Process quality was never a knowledge problem. It was a visibility problem. Ramblr Inspect gives BASF's teams the evidence, guidance, and institutional memory to act before failures occur – and to get better with every installation.

• Failure points identified and documented before they cause downtime or safety incidents

• Error-prone installations automatically flagged – QC effort directed where it's needed most

• QC teams guided through inspections with AI-verified video evidence

• Frontline operator knowledge systematically captured and fed back into training

• Consistent quality standards maintained across plants, teams, and experience levels

• Scalable foundation – one AI system, deployable across sites and processes