Department of Science and Technology
An AI assembly trainer and digital twin for manufacturers (DMAT-DTS)
A DOST-funded Digital Manufacturing Assembly Trainer and Digital Twin System that helps manufacturers train assembly staff, now approved, in classroom testing at DLSU, with two companies lined up for deployment.
Who it's forManufacturers that need to train assembly staff faster and more safely, without tying up the production line.
- DOST-funded project
- Approved
- companies waiting for deployment and testing
- 2
- classroom trials at De La Salle University
- In testing
The challenge
Training assembly workers on a real production line is slow and expensive: it ties up equipment, takes experienced staff away from production, and mistakes are costly. General-purpose AI chatbots don’t solve this, because answers about assembly must come from the company’s approved procedures, not from the open internet.
What we built
The Digital Manufacturing Assembly Trainer and Digital Twin System (DMAT-DTS):
- A digital twin of the assembly process, so trainees can learn and practice without stopping a real line.
- An AI trainer that answers questions only from approved manuals and procedures, using hybrid keyword and semantic search so it finds the exact right step.
- Built-in evaluation, so answer quality is measured rather than assumed.
Where it stands
The project is approved and DOST-funded. It is being tested in classroom settings at De La Salle University, and two companies are waiting for deployment and testing on their own lines.
