Context
World leader in inspection: every year, Bureau Veritas carries out thousands of inspections of industrial components, essential for guaranteeing the conformity and safety of installations.
Time-consuming process: each visit takes several hours of an inspector’s time, between manual identification of equipment and inputting reports.
Need
- Improve speed: reduce the time spent on each inspection to free up expert time.
- Improve reliability: reduce input and identification errors by automating component recognition.
- Add value to data: transform snapshots and reports into a structured database to feed other cross-functional services (predictive analysis, quality monitoring, maintenance).
- Working offline: the solution needs to be turned directly on to mobiles, as some inspection sites are in white zones with no network access.
Completed work
Development of an application for taking photographs of the elements to be inspected, automatic detection of devices and possible correction of results by inspectors.
- Expert partnership: Datalchemy and Micropole have combined their AI and mobile application know-how to meet the needs of Bureau Veritas.
- Offline mobile application:
- Direct installation on the inspector’s smartphone, with no network connection required.
- Intuitive interface: photo capture, annotation and validation in just a few clicks.
- Automated inspection:
- Automatic recognition of the elements to be detected
- Instant label assignment and pre-filled report generation.
- Embedded AI on cell phones.
Results
- Detection performance: over 80% of components automatically identified.
- Time savings: significant reduction in average inspection time.
- Sustainable capitalization: structured data collection ready to feed new analysis, predictive maintenance and quality management services.
This innovative solution is a concrete illustration of the contribution of AI at the heart of the inspection business: faster, more reliable and generating value for the entire organization.