AI for Context Recognition

Context

The construction sector presents complex operational scenarios, characterized by the simultaneous presence of vehicles, equipment, and constantly moving safety barriers. In this context, monitoring construction site activities and safety conditions is particularly challenging, especially in environments subject to environmental variables.

Solution

The Proof of Concept developed consists of creating an AI model based on instance segmentation algorithms for the automatic recognition of various construction vehicles and types of safety barriers.
The system operates in real-time on images captured by the camera and is capable of accurately identifying objects in complex and dynamic scenarios. The developed architecture ensures high accuracy and easy integration with autonomous mobile platforms, thus providing a solid foundation for monitoring and safety applications in construction sites.

Impact

The project demonstrated the technical feasibility of an advanced system for the automatic recognition of vehicles and barriers in construction sites through the use of instance segmentation. The entire process highlighted a strong ability to adapt to the real operational needs of the construction sector. The choice of the final model ensured high accuracy, allowing precise recognition even in complex scenarios. A scalable system ready to be further optimized and integrated into production contexts.

Funded by the European Union – Next Generation EU

Client

Monitor the planet

Sector

Services

Technological area

Digital Factory

Technologies

Artificial Intelligence | Computer Vision | Machine Vision

Field Expertise

Context analysis and fine-tuning, data collection and analysis, laboratory experimentation of AI-based models

Follow us on LinkedIn®

Don’t miss any updates on tenders, job openings, and CIM news directly in your feed.

Schedule an appointment with one of our experts

"*" indicates required fields

This field is for validation purposes and should be left unchanged.
Personal Information*