Amount awarded: £1,200,000
Led by: Anglian Water Services
Partners: Avanade UK Limited, Microsoft, Severn Trent Water, Wessex Water, National Trust, The Country Land and Business Association (CLA), Suffolk Sustainability institute – University of Suffolk, Norfolk Rivers Trust
Project completed: November 2023
This project brought together Anglian Water and a partnership of organisations to explore how digital twin and AI technology could be extended to the natural world to address the problems we collectively face across the water sector and wider society in a way that could be scaled to all water companies across the UK. This project aimed to deliver short and long-term outcomes aligned with Ofwat’s strategic innovation themes. Firstly, to directly improve the health of our rivers, overall environment and the general public’s perception of river water quality. The platform aimed to provide a novel way to engage all stakeholders in the health and ecological management of UK rivers, and to deliver directly or indirectly, against Ofwat’s strategic innovation themes. The ecological digital twin aimed to help protect and enhance the future natural environment by providing a new capability to manage and model ecological health and the impact of the whole stakeholder community.
As part of Anglian Water’s purpose to create social and environmental prosperity for the region, in 2022 Anglian Water joined forces with Severn Trent to Get River Positive by 2030. Working collaboratively, Get River Positive is driving real action and making the changes we all want to see for a better future for our rivers and environment through a wide variety of partnership-led projects, including this one, which aligned perfectly with the Get River Positive commitments.
Together, the partnership explored the problems we collectively face across the water sector and wider society; Only 14% of rivers in England are at Good Ecological Status, 10% of freshwater and wetland species are threatened with extinction in the UK, and we’ve lost 90% of our wetland habitats in the last 100 years. Healthy wetlands and rivers provide communities with protection from flooding, access and amenity and physical & mental health benefits. Many sectors impact river health and there are gaps in system thinking leading to sub-optimal outcomes. Our goal for the project was to accurately understand the health of the river system and the positive and negative impact of Anglian Water’s and other operations to enable effective action
| Partner | Role |
| Microsoft and Avanade | Technology partners |
| Norfolk Rivers Trust, National Trust and the Environment Agency | Environmental and social value partners |
| Country Land and Business Association (CLA) | Agriculture and landowner representatives |
| University of Suffolk | Research and academic partners |
| Severn Trent and Wessex Water | Wider water industry representatives |
A Digital Twin is a digital model of a real-world entity, in this case, a river, which enables the digital application of different inputs and scenarios in order to view the impact of these within the safety of the digital environment. With our technology partners, we began to create ideas of how extending digital twin and AI technology to the natural world could address the problems we faced and achieve our goal. We settled on creating a world-first River Quality Digital Twin, initially focused on the River Stiffkey, one of Norfolk’s iconic chalk streams, but to be built and developed in a way that it could be applicable and scalable to any water company and river system.
Our theory was to first ‘understand’ the complex system of a river by mapping it within a common framework. We captured this in a River Health Ontology, which we could recreate as a digital model in a Digital Twin. The ontology defines the categories and properties of a river and how they relate to each other. It is applicable to all river systems and can be widely shared. With this digital representation of a river system, we then collated disparate datasets, aligned to the River Health Ontology model, into a single digital twin environment and connected via Anglian Water’s existing data platform. With the River Quality Digital Twin being fed by data, we have opportunities to carry out complex system modelling of assets, processes and resources, to provide machine-led diagnostics and embedded analytics engines. This will provide future opportunities to determine cause and effect. With this clarity, there is potential for different organisations to come together to work collaboratively to determine the best course of action to protect and improve river health.
In January 2023, an initial prototype solution was jointly funded by Anglian Water, Microsoft and Avanade over a 3-month project. Using expert knowledge from across the partnership, the river system ontology was created in an Azure Digital Twin. The Microsoft Azure platform helped to accelerate this as it provided us the tools and technology in a completely connected stack to implement ideas and test them quickly, utilising technology already used by Anglian Water and widely across the sector.
Alongside this, data sources were reviewed with partners for input into the prototype. Twenty-eight data sources were identified ranging from discharge flows, to weather, to river geojson files, to cropping and soil data. A data ingestion pipeline was built to store data in a data lake, transform it and ingest it into the Azure Digital Twin (ADT). With the digital representation of the river system populated with existing data, a UI layer was created in PowerBI to show trending and a spatial map of the river Stiffkey.
In May 2023 Ofwat Innovation funding was secured to continue developing the River Quality Digital Twin and move our prototype to an MVP stage.
The goal of the Ofwat Innovation Funded Project was to focus on use cases for the digital twin and to create a data science model that would provide an option to ‘look forward’, as well as onboarding more data. Three products were created, with varying levels of maturity:
These products are distinct in terms of their usage to meet varying objectives, but all use the underlying River Quality Digital Twin capability that had been developed during the prototype. The River Quality Digital Twin capability was enhanced by increasing the data and coverage of the ontology, including installing new IOT sensors into sections of the river system to gather more data. A Structural Equation Model (SEM) was created to capture the state of the river as a series of observed values and equations linking each of these values. Our SEM captures the relationships between different datapoints on the river. This enables a view of the downstream impacts on the river of making a given intervention.
We have developed a River Quality Digital Twin of one of Norfolk’s iconic chalk streams, the River Stiffkey. Seamlessly integrating the digital twin with historical data and data science models has provided Digital Cognition capability to Anglian Water, bringing clarity to Stiffkey’s complex system, providing insight which will turbo charge the response to river water quality to meet the evolving needs of customers, wider society and the environment. The solution can be scaled to positively affect customers and the environment now and in the long term, enabling new ideas and ways of working with partners from a broad range of sectors and perspectives. Taking an open approach, our project will generate learning for the benefit of all. We have created a step-by-step guide to support other companies in adopting the River Quality Digital Twin, in the spirit of the Ofwat Innovation fund and supporting knowledge sharing across the industry.
—
For more information on The World’s First Digital Ecological Twin, take a look at the following resources:
