Make flotation observations measurable
A leading mining and metallurgical company in the CIS / Asia region evaluated computer vision for visual monitoring of flotation machines. The project was delivered as an MVP demonstrating the technology’s capabilities.
Parameters extracted from video
- The type and pattern of froth movement.
- Operation of the machine’s visible mechanical components.
- Bubble size and bubble count.
- The speed at which froth is removed.
The system processed imagery of the flotation machine’s working area and produced structured observations for process specialists. This changes the output from a video to watch into parameters that can be reviewed and compared.
Documented result and project stage
The MVP confirmed the feasibility of automatic flotation-froth analysis and the generation of structured data for technologists. It is presented as a technology demonstration, not as a completed plant-wide process-control deployment.
Engineering the next stage
A production rollout would need stable camera geometry and suitable lighting over the froth, representative recordings across operating conditions, and comparison with process specialists’ observations. The useful parameter definitions, sampling intervals and connection to process data should be agreed before integration.