OMNI is a bespoke machine learning system that helps both vehicle parts suppliers and vehicle assembly lines to improve key plant metrics, and to reduce downtime, rework, and scrap. OMNI achieves this without additional hardware requirements, by using only data that you already own.

Today’s vehicle manufacturers rely upon Just In Time (JIT) production systems that are lean by definition. These are typically characterised by:

  • A critical need to preserve supply chain logistics,

  • No tolerance for defects, errors, or repairs, and

  • The use of performance penalties to recover from the impact of a defect on these inflexible lines.


Based on historical production data, OMNI accurately predicts the presence of defects. OMNI does so by recognising patterns in the materials, parameters, and conditions that have previously been associated with one or more known defects. In this way, OMNI can detect defects that you would not otherwise be directly aware of, such as:

  • Subsurface defects
  • Latent defects (particularly important where warranty claims could be significant)
  • Uncaught quality violations
  • Hidden scrap

Traditionally, the cost of detecting these types of defects has been high, in cases incurring:
  • Shipping
  • Factory downtime
  • Disrupted logistics
  • Customer dissatisfaction

OMNI helps manufacturers to find and quarantine bad components before they become costly.

Parameter Optimisation

In addition, OMNI will optimise your manufacturing process, thereby preventing defects from occurring in the first place. The AI at the heart of OMNI learns and then prescribes the combinations of materials, process parameters and other conditions that result in a higher yield.

In this way OMNI steers the manufacturing process into an operating region that:

  • Lowers the cost of non-quality
  • Increases efficiency
  • Provides higher yield

OMNI Stud Welding Case Study

OMNI Stud Welding Case Study

75% reduction in stud welding defects.

OMNI Manufacturing Case Study

Predicting engine block defects and identifying high yield operating regions.