Physical services
Overview
Emphasis DigiWorld, a developer of sensor-based industrial intelligence systems from Greece, worked with AI-MATTERS, the Manufacturing sectorial TEF, to build an AI-powered solution for assessing concrete workability during mixing and transport. The collaboration combined AI-MATTERS' technical guidance, network of domain experts, and structured testing support with Emphasis DigiWorld's sensor hardware and IoT engineering capabilities, resulting in a deployable system now being trialled on active construction sites. John Stivaros, IoT R&D Engineer at Emphasis DigiWorld, led the project on the company side.
Impact
The resulting system predicts concrete workability with 90% or higher accuracy (R² = 0.85) using real-time sensor data, while running at low latency and minimal CPU load — light enough to deploy on embedded devices. The hardware itself is designed for rapid, non-intrusive installation: a wireless device that attaches magnetically to mixing trucks, requiring no changes to existing equipment. "The results we've achieved with AI-MATTERS exceeded our expectations. It's not just about the technology — it's about transforming how we think about quality control and automation in heavy industry," Stivaros adds.
Emphasis DigiWorld now plans to extend the solution to additional construction sites and explore further AI applications for real-time quality control across other materials and processes. "AI-MATTERS has been a true partner in innovation, helping us navigate the complexities of AI adoption while driving real results," Stivaros concludes.
The Challenge
Concrete workability — how easily a mix can be handled, poured and finished — is traditionally assessed through manual slump tests, which require operators to interrupt production and interpret results by eye. Emphasis DigiWorld set out to replace this with a system capable of predicting workability in real time, directly from sensor data, without operator intervention. Doing so meant closing the gap between raw sensor readings and reliable, actionable predictions — a data engineering and model validation problem as much as a hardware one. "We knew AI could transform how workability is assessed, but bridging the gap between raw sensor data and actionable insights was a challenge," says Stivaros.
The Solution
AI-MATTERS supported Emphasis DigiWorld through collaborative workshops, tailored technical guidance, and access to domain experts who understood both the AI and the construction materials side of the problem. This combination allowed the team to engineer higher-quality training datasets, iterate on machine learning models, and test the system under realistic industrial conditions rather than in a lab setting alone.
