Digital twin technology is rapidly making its way through the telecommunications sector. Digital models of physical objects, systems and processes are being increasingly used for real-time remote monitoring, predictive maintenance, risk management and troubleshooting.
Sparkbit, a member of the PrediqTank consortium, is working with a telecommunications startup that creates photogrammetry-based 3D models of cell towers to automate the detection and analysis of digital twins. The overall goal is to provide a comprehensive solution for monitoring and maintenance of cell towers and masts.
Our team is currently developing machine learning (ML) algorithms to extract data from 3D models of cell towers to automatically identify and assess resources and assets installed on the towers, as well as evaluate the utility of the physical space available on the towers. The main benefit of this approach is that the manual efforts required to prepare fully analyzed digital twins, which otherwise take weeks to accomplish, are now significantly reduced by the factors of 100-1000, enabling much higher scalability of the process.
Project Category: ML
Lead Member: Sparkbit
Technology used: Python, Tensorflow, Keras, Scikit learn, C++, Terraform, AWS