Retrocausal

Retrocausal is a manufacturing process management software provider whose AI "Copilot" products apply patented computer vision and machine learning to manual assembly operations. Its Assembly Copilot uses off-the-shelf or overhead cameras at workstations to monitor assembly steps, deliver AI-tracked digital work instructions and poka-yoke step verification, flag operator mistakes through visual and auditory cues, and record cycle-level video for traceability and recall investigations. Aggregated workstation data feeds analytics dashboards covering cycle time, yield and error rates, supporting root-cause analysis and identification of training needs.

A second product line, Kaizen Copilot, targets industrial engineers and continuous-improvement teams and runs from smartphone or web-camera video. It automates time and motion studies (including MODAPTS analysis), line balancing, ergonomics assessments with recommended hierarchy of controls, standard work definition, and quality planning artifacts such as P-FMEA and control plans. The company states it has developed multi-modal foundation models for manufacturing trained on proprietary datasets, has been granted three US patents on its generative AI technology with four more pending, and offers pre-built integrations with common MES systems, knowledge bases, smart tools and IIoT infrastructure. The platform runs on Amazon Web Services, is ISO 27001 certified, supports Azure Active Directory, Okta and IBM QRadar, and includes worker-privacy features such as facial blurring and fixed regional pixelation.

Retrocausal markets vertical solutions for segments including electronics, automotive and automotive parts, medical devices and HVAC systems manufacturing, and publishes customer case studies involving Siemens, 3M, a water heater manufacturer and a lighting control panel manufacturer.