Lessons from Anglian Water’s large-scale Digital Twin deployment
As a company expands, and so does its IoT applications, the vast data it collects to give insights, could begin to become a blur. Although things like AI can condense and streamline mass quantities of it into actionable insights, understanding the knock on effects remain harder to diagnose, even for an AI model, should relevant data remain unfilled.
No better example of this scenario is the water utilities sector. A complex network of water pipes spreads out like a spider’s web under our towns and cities, these pipes run below streets, houses, and public buildings, forming an extensive system that carries water to homes and businesses. This network criss-crosses under the urban landscape, connecting different areas and ensuring that water is readily available wherever it’s needed. So, imagine the impact of this if something goes wrong. Firstly, diagnosing where exactly in the network this is happening would prove difficult, as the records of such pipes in the UK are either dated or non-existent. Equally, trying to ascertain what knock-on effects an issue in one part of the pipe system could have further along in the system means companies are often left on the backfoot to be reactive to problems instead of proactive.
Issues like this is why Anglian Water, the largest water and sewerage company in England and Wales, is leading a project called Safe Smart Systems. “We’re targeting the Safe Smart Systems Digital Twin to address one of the most fundamental challenges for any water company right now – how to keep the taps running as climate change affects water supplies,” Fionn Boyle, Strategic Innovation Lead at Anglian Water told IoT Insider. “Needing to run and build a more resilient water infrastructure demands a much more innovative system and data-led approach to make the right operational decisions. This is where a Digital Twin can play a pivotal role by providing water management with a much fuller insight into the dependency of their data.”
Anglian Water’s project implements the use of AI and mathematical optimisation to improve long-term operational resilience of water systems across the UK. While their focus has been on the management of basic utilities, their development and deployment of such a project has a lot of perspective to offer for many industrial players.
Digital Twins from the ground up
Digital Twins are advanced technological models that create a virtual replica of a physical system, process, or product. These digital counterparts use real-time data, simulations, and analytics to mirror and predict the behaviour of their physical counterparts. By integrating data from various sources – such as sensors and IoT devices – Digital Twins enable users to monitor, analyse, and optimise the performance and operation of the physical entity they represent – serving as a bridge between the physical and digital worlds, providing deep insights into the workings of complex systems, and enabling proactive decision-making.
How this works in Anglian Water’s case is, the Digital Twin for Safe Smart Systems will stream live data about water pressure and flow from anywhere in the water distribution system. They will be able to spot what changes are happening and whether these are normal or abnormal readings that need to be investigated. The Digital Twin then allows them to run through thousands of scenarios in real time to understand what the optimal response to that change in the system is and how they can resolve it in the best way possible.
“Over time we expect the capabilities of Smart Safe Systems to grow to a point where we will have the ability to automatically deploy a response to a water supply issue without the need for a human intervention,” explains Boyle.
This automated response could come through its use of AI and machine learning seeing the data and automatically implement corrective measures, such as adjusting valves or rerouting water flow. This level of automation is possible because the Digital Twin has a comprehensive, constantly updated model of the physical system, allowing it to simulate and predict outcomes with high accuracy.
Developing a Digital Twin
Just how Boyle has explained future plans to develop automated responses to supply, the development to get Smart Safe Systems to where it is today has required a herculean effort.
The inception of Safe Smart Systems at Anglian Water marked the beginning of an ambitious journey to digitally mirror their water distribution network. Initially, the challenge was twofold: understanding where automation and analytics could yield the best outcomes and integrating the Digital Twin with the complex, sensor-laden infrastructure. A significant learning curve was adapting to the real-time data requirements, a departure from traditional, static models. This required rethinking not just the Digital Twin’s construction but the entire supporting environment.
Historically, Anglian Water’s models have been built and used for a point in time, such as building a new storage tank. But with a Digital Twin like Safe Smart Systems, models need to be running continuously and updating until you find the best optimal solution. This a whole different world in terms of the flow of data through, meaning thought is having to be put into not just building the Digital Twin, but how to build the environment that then supports the operation and testing of that solution to ensure that minimal faults in it in its operation. Challenges like these presented a hurdle for Anglian Water in implementing their Digital Twin system.
“The implications of being a legacy industry extends to the state of our data quality and how that causes issues with how our operational technology runs,” said Boyle. “In transitioning to the Digital Twin system, we knew data quality stood in our way especially when we were thinking about applying machine learning and AI.”
Strategic insights
One of the critical lessons from Anglian Water’s experience is the importance of scalability and organisational readiness. While Safe Smart Systems began as a localised project, it was designed with the potential for wider regional or even national deployment.
“Even when the scenario is so big and complex, you need to start with a user-based approach and work out what the specific issues are for your business,” says Boyle. “Don’t be distracted by the shiny technology and understand the Digital Twin is going to be something built from the bottom up with a lot of effort focused on knowing and improving existing data structures.”
Documenting the journey, from design through implementation, has created a valuable playbook for scaling up future Digital Twins. A key consideration in scaling up is understanding the sequencing from physical assets to data fusion and analytics. This process requires a nuanced understanding of the project’s maturity and the gaps that need bridging to achieve the desired capability level.
The integration and management of diverse data sources have been pivotal in the development of Safe Smart Systems. The transition to a Digital Twin system highlighted the importance of data quality, particularly when applying machine learning and AI. “A large part of the project so far has been digging into our data, which is so fundamental for the success of the Digital Twin. Tidying up our data sets is what will have the longest lasting impact, rather than just the AI decision engine that sits on top,” explains Boyle. This focus on data integrity and structure is a critical lesson for any industry looking to implement Digital Twin technology.
Anglian Water’s approach with technology partners has also proved instrumental in the development of Safe Smart Systems. “We have twenty-seven partners in total, giving us access to skilled and knowledgeable people from the water and technology industries, suppliers, and academia,” says Boyle. “We have since continued to include our partners during every step of the project. It is a continuous reminder to ensure that our system is scalable and not built too restrictively around the Anglian region.”
The involvement of a wide array of partners – from industry experts to academics – has fostered an environment of knowledge sharing and innovation. This collaborative process ensures that the Digital Twin is not just tailored to Anglian Water’s current needs but has the flexibility to adapt to different geographical and operational contexts if it expands.
Yet regardless of all the data and tech behind the pursuit for such a project, Boyle highlights the human element of it all: “Make sure you have the right change management in place early on and that you’re building the right culture across your organisation to accept the technology.”
Implementing a Digital Twin is preparing the organisation for the profound changes it brings. Anglian Water’s experience underscores the need for effective change management, communication, and storytelling. The shift from a traditionally reactive business model to a proactive one necessitates a re-examination of team structures, collaboration methods, and incentive models.
Although full results are anticipated in the latter stages of the project, Boyle expects Digital Twins to revolutionise predictive maintenance and asset management strategies: “It is early days, but the first benefits of the Safe Smart Systems Digital Twin will be transforming our capability to do proactive maintenance by predicting when and where repairs need to be done ahead of time. The next stage should be an instant response capability by using AI and machine learning to automate how we spot a leak or a water quality event. As we gain confidence, it is possible for our Digital Twin to do something that no other Digital Twin used by a utility has done before. That’s raising the level of autonomous operations to the next level and support self-healing water supply networks.”
It is for reasons like this, that Boyle predicts Digital Twins will soon become integrally intwined to how infrastructure businesses operate. For a utilities company, that carries its own challenges. But like with Anglian Water, the Digital Twins implementation is forcing them to rethink how they structure themselves; not just in terms of hierarchy, but capability-wise and the skills the organisation will soon have greater need of.
“The skillset needed to work in our industry is going to change dramatically in the future and this is something we must be prepared for,” asserts Boyle. “Whereas today you could argue that the water industry isn’t a typically digital native industry, over time it will.”
Deploying Digital Twins
Anglian Water’s journey with Safe Smart Systems offers a blueprint for other companies venturing or pondering large-scale Digital Twin rollouts in their operations. The advancement Safe Smart Systems presents in its real-time monitoring and response mechanism of a large network of connected devices speaks to the application this can serve in other industrial operations.
The takeaways their project highlights are the significance of thorough planning, data management, collaborative innovation, and cultural readiness to be able to adopt and develop one for your unique operations.
As Anglian Water testified, there are upfront costs and challenges associated with deploying a digital twin to your operations. Yet, the future of infrastructure management, as Boyle believes, is intertwined with the technology, and those who take it head on will get a better grip on it.
The benefits of more cohesive Infrastructure or asset management will duly help return the down payments associated with its development and deployment as the new insights yield greater foresight and preparedness of disruptions, reducing costs and downtimes associated with the issues. This plus the booming use and application of AI means the technology, as Boyle explained, means those with the Digital Twin infrastructure deployed could soon be closer to seeing self-healing networks, a concept that could redefine infrastructure management across industries.
There’s plenty of other editorial on our sister site, Electronic Specifier! Or you can always join in the conversation by commenting below or visiting our LinkedIn page.
