BUILD-DT

Challenge #4 - Open challenge

BUILD-DT

Scalable and Resilient Construction Site Digital Twin in the Computing Continuum

CHALLENGE #4

Open challenge

Enhancing the scalability and resilience of a construction site digital twin with NebulOuS

Description

BUILD-DT tackles the problem of intelligent health and safety monitoring, striving to detect worker health hazards as fast as possible. To do so, it utilizes a digital twin system. The system is continuously updated with real-time field analytics derived from IoT sources (camera, wearables, weather sensors). BUILD-DT uses the analytics in three specific hazard detection pipelines. 

The first detects workers that enter the site without proper PPE (personal protective equipment). To do so, it performs AI analysis on stream video data from cameras mounted at the entrance of the site. The second utilizes the location and acceleration data from UWB tags placed on workers to detect worker falls. The third combines the data from an on-site weather station and worker’s wearables to determine whether a person’s health is at risk of overheating or UV overexposure. Each of the pipelines produce a notification sent to the OSH manager and, in case of worker health monitoring, the worker in question, as soon as an incident is detected.

ENTITIES

NeverBlink

(Poland) Deep-tech startup developing streaming knowledge graph and neurosymbolic technologies

Team Members

Karolina Bogacka

NeverBlink’s CEO and co-founder. Efficient edge AI expert.

Piotr Sowiński

NeverBlink’s CTO and co-founder. Knowledge graph expert.

Anastasiya Danilenka

NeverBlink’s lead data scientist.