Triton — a digital twin for water distribution
At DataThings we specialize at creating efficient digital twins with tangible business value for our clients. Grasping the potential benefits of mapping a system from the real, physical world, to the digital realm can be tricky though. Understanding existing examples often requires very specific domain knowledge, and presenting them publicly is usually hindered by non-disclosure agreements or data protection regulations.
Fortunately, Triton hits just the right level of complexity to allow us to highlight many cool features of digital twins in general and our Greycat technology in particular. Thanks to our partner in developing Triton, DEA (Distribution d’Eau des Ardennes), we can even use real sensory data to do so. In the upcoming weeks we will publish a series of short blog posts, each focusing on one specific aspect of a digital twin, made by DataThings, and how we could help you innovate your business use case.
Introducing Triton
As its divine namesake might suggest, Triton is all about water. More specifically, Triton is the name of the project which tackled the development of a digital twin of the water distribution network of the north of Luxembourg, governed by our partner DEA. The project started in 2023 following a call for tender from DEA in the context of the AI4GOV initiative. For simplicity, we will use the name Triton also to refer to the application that has been developed for the Triton project.
Triton consolidates several data sources and types of data that DEA need to handle in their daily business. This includes thousands of time series form sensory data, geo-data to determine the location of network elements, but also annotations provided as csv files. Triton provides easy and fast access to live data from a map-based web interface and scans the sensory data for anomalies. The latter allows Triton to inform DEA employees about potential leaks in their network, a crucial feature to reduce tedious manual work on DEA’s side.
The physical counterpart
The mission of DEA is to manage and distribute the portable water to 28 municipalities in the northern part of Luxembourg. Here are some numbers to outline the scope the network that is covered by Triton:
- 3 groundwater sources
- 15 pumping stations
- 34 regional tanks
- 160 local tanks
- 492 manholes
- 1.6k pipe elements (466 km)
- 8k time series from sensors
The median time span between two values on the time series is 10s, meaning that Triton tracks and handles 2.8 million values per hour. All of this live data is accessible with a few clicks and regularly checked for anomalies without any input or disturbance to the application users.
There is no considerable coverage of smart meters at household in the north of Luxembourg, yet. Therefore, the network currently ends at the local tanks that redistribute the water to the end-consumers. From a technical standpoint though, Triton powered by Greycat is ready to be extended when smart meters become more widely available.
Wrap-up
This article introduced Triton, a digital twin for water distribution and leak detection. We have covered the origin of Triton and hope that we have invoked your interest! If so, stay tuned for more posts about Triton in the coming weeks. We plan to cover the whole pipeline in individual articles, including data modelling and visualization, user interface and experience as well as anomaly detection. But before that, please find here below a sneak peek at Triton.
If you are interested in testing out GreyCat, we offer our community version for free and provide documentation for our GreyCat language as well.
Articles in the Triton series:
- Triton — a digital twin for water distribution (this article)
- Modelling a digital twin for water in GreyCat
- On the importance of intuitive user interfaces for digital twins
