A recruitment consultant will contact you soon.

AI Visual System

The Energy sector is undergoing a major transformation

Innovative AI system built to train in defect recognition for industrial QA using image recognition and deep learning.

Context and Challenges

The AI Visual System involves the collaborative work of Infotel developers , AI data scientists, business managers and QA experts to capture and refine banks of images for training in the automated recognition of defects from dents & holes to overpainting & missing elements. The software has been developed to allow for adaptation to pretty much any industrial process, building in variables to:
• fine-tune to different environments and station sensors
• encompass a range of defect types
• apply alternative deep learning models
• enable the future potential to extend from object detection to
anomaly detection and optical character definition



Goal of Project & Key Success Factors

  • GOAL OF 98% ACCURACY for recognition of defects when deep learning training has reached a valid plateau
  • ACHIEVING HIGH 90s ACCURACY depending on defect task and environment (98% goal on clearer subjects such as holes and missing badges or where conditions are more easily optimized to reduce false positives e.g. low light conditions).
  • Computer vs humans: HUMANS ACHIEVE LOW 90s ACCURACY.
  • Our client have actioned plans to roll out the AI Visual System factory wide , encompassing a range of QA processes in different environments.


Project challenges and enhancements

The major challenge / innovation has been in refining to adapt to real industrial environments , ironing out false positives and environment variables, with the objective of producing a stable, scalable and performant system. Challenges include working within processing parameters (power required for images is huge, optimisation key); generating sufficient test samples for training.

Software currently being adapted to enhance visuals to show where defects are on vehicle (can be difficult to spot exactly where a dent is on otherwise featureless panels).

Bottleneck in image capture to data pre processing , to be addressed with plans for point and click automation.





infotel logo

Words from an Infotel developer

The focus has always been to industrialise the proof of concept, into a stable, scalable, performant and rapidly deploy able factory wide solution

Solutions and Actions


• Adapt and fit industrial environment factory stations) for QA monitoring and provisioning of test images
• Build for creation , training and refinement of defect models
• Apply automation process to live environment


• Automate QA to detect and flag defects to 98% accuracy
• Ability to establish and train new defect models
• Ability to adapt to different environments
• Industrialize and deploy factory wide

Identify rejects / remediation requirements early in the manufacture process to reduce costs.


  • 1 Project Manager
  • 2 AI Data Scientists
  • 1 Business Lead
  • 6 Business Pilots
  • 1 Front End/Back End Developer


The stack

  • Microsoft Visual Software
  • Microsoft SQL Server (DB engine)
  • C# coding
  • Google Tensorflow (open HTML CSS Vanilla JS)
  • Mosquitto (MQTT) transmission protocol transmits between
    separate layers , providing a lightweight method of carrying out
    messaging using a publish/subscribe model


Automatic process of loan request and the simplification of KYC "Know Your Customer".


  • Improvement of KYC and the onboarding of new clients with automatic identity verification thanks to intelligent ID document processing on the web or in an application.
  • Acceleration in the dealing with requests for loans and mortgages: process and automatic classification of new documents, extraction of information and validation of data points.
  • Fraud prevention and respect of rules: automation of compliance requirements on contracts, financial documents and ID documentation.
  • Easing and simplification of the consumer spending comprehension: classification of bank transfers from the study of receipt and invoice item data.


Automated process of scientific papers and other documentation linked to the registering of a patent for intellectual property.


  • Recognition and automatic indexation of fields for a precise search among the documents relating to the patent application.
  • Standardisation of the documents with online specifications: bibliopraphy, expired patent, renewed patent, etc.


Automatic processing of receipts, bills and other financial documents.


  • Automisation of the accounting process: declaration of VAT, supplier bills, etc.
  • Standardisation and harmonisation of the business accounting processes.
  • Improving data security with the deletion of manual errors and with the synthesising of controls.
  • Contribution to the reduction in fraud risks, à la réduction des risques de fraude, reinforcement of operational security, audit trails and internal controls.
  • Support in the change management and accounting teams on the appropriation of this new solution.