Skip to main content
Home | What we do | Use cases
BI-REX

Predictive quality control and fault detection in industrial 3D printing

Reducing waste and optimising materials in advanced manufacturing through computer vision and real-time AI orchestration.

Markets
Solutions

The challenge

Metal additive manufacturing faces serious challenges around process stability and production costs. During the printing cycle of critical components, micro-defects or anomalies in individual material layers can compromise the structural integrity of the whole part. Without continuous, automated monitoring, non-conformities are only caught once the job is finished, leading to a significant waste of precious metal materials, time and production capacity.

The challenge lies in the need to support the human operator in real time: anomalies must be intercepted layer by layer (slice) by processing continuous image streams directly on the machine, turning quality control from traditionally ex-post inspection into predictive, immediate governance of the industrial process.

How we create value

Optit met this challenge by developing a custom software application that acts as the connective tissue between industrial hardware and data science. At its core, the value lies in a logical architecture designed to manage massive real-time data flows without latency.

This solution demonstrates Optit’s flexibility in integrating natively with proprietary AI and fault detection modules, such as the computer vision models developed by Bi-Rex.

The orchestrator captures images directly from the 3D printer, securely routes them to the AI model and instantly translates its output into clear operational guidance. Through an interactive dashboard, the system maps anomalies layer by layer.

In this way, the application turns algorithmic complexity into a concrete governance tool that supports the operator in critical decisions to halt or correct production.

Impact

Introducing the fault detection platform delivered immediate operational benefits:

  • Real-time anomaly detection and visual, layer-by-layer monitoring of the print job.
  • Reduced industrial waste, directly saving materials, energy and machine hours.
  • Enhanced decision support for the operator through centralised process metrics and KPIs.
  • Modular, scalable architecture ready to integrate future performance management or production planning services.
Predictive quality control and fault detection in industrial 3D printing