Renault had some assembly problems leading to re-work and wasting time. This was mainly occurring due to the volume of vehicle types in the same line. Therefore they requested an image recognition solution to solve this.


Our team in LATAM didn't have experience in the Biovia Pipeline solution. Still, we knew it could run machine-learning scripts; then, I was challenged by Henry Shibayama, Diogo Faccio, and Vinicius Pinto to structure and develop with then a solution in my preferred programming language to solve Renault's problem. 

I decided to do it using python due to my prior experience with it.

The POC developed was an Assembly Assistant, which notified the operator if the assembly sequence was right and if all steps were executed.

Customer On-Site Pilot

After the concept being proved, we started working with Anand Krishnamurthy, Biovia Pipeline Pilot Expert, to develop a pilot inside Renault Factory. Henry and Diogo led this project and presented it.

POC Presentation | Demonstration

Download Concept Presentation:


I used OpenCV in Python 2.7. The method utilized Haarcascades to detect the objects. 

Download Project Detection Presentation:

Code Flowchart