0 of 0 for ""


Reading time: 7min

Craig Hardingham

Digital Technical Director

AI in water management: how digitally future-proofing infrastructure can produce major efficiency gains


The water and wastewater industries are under pressure to refine their operations and upgrade infrastructure, but the complexity of the design and engineering process means any process inefficiency becomes costly and time-consuming. Here, we look at how Artificial Intelligence (AI) tools hold the key to optimising decision-making.

Safe and reliable water and wastewater services are critical, both for protecting public wellbeing and safeguarding the environment.

Embedded, quite literally, into the communities they serve, water companies have a big opportunity to generate a positive impact for both residents and the wider environment. But with this opportunity comes pressure to respond to the growing challenges of climate change and population growth, and adapt to evolving customer needs, all while ensuring that services remain affordable and accessible.

In its 2019 strategy, Ofwat highlighted that a solution must be for water companies to embrace innovation, including new technology, to future proof their key infrastructure. Doing so will not only enhance performance, but benefit customers, communities, and the environment.

Efficient design

A key consideration for water companies, however, is ensuring that any investment is cost-effective. Our experience in the water, energy & industry sphere has shown that often, balancing the anticipated cost with available investment can lead to delays in realising the benefit.

The challenge for the water engineering industry is to ensure that they can offer workable solutions, able to satisfy the often-complex requirements of water and wastewater infrastructure projects, that champion efficiency with minimal delays to the investment cycle.

The most critical project stage for this is upfront design. The list of constraints that any new piece of water infrastructure must comply with is extensive, and incorporates everything from physical geometry and site conditions, planning or regulatory restrictions – especially around environmental impacts – to requirements of landowners and other stakeholders.

This inherent complexity means there are numerous potential solutions, all offering differing sets of benefits and disadvantages. The design process naturally involves making a huge number of significant decisions, all of which can alter a project’s direction of travel – often multiple times within these earliest stages.

Knock-on effects

Working in the traditional way, the process of exploring and understanding the various outcomes of decisions made at this critical stage is enormously time-intensive, if not impossible. Often, decisions are based on intuition and the experience of the individual designer.

This is where AI can play a crucial role, by modelling these various potential outcomes and allowing designers to understand the ramifications of any given decision instantly.

Typically, when changes need to be made to a design that has already been detailed, they can have knock-on effects throughout the whole design. But with AI, designers can quickly check the feasibility of these changes and implement them, where traditional ways of working may have required a significant amount of redesign and inherent delay.

Using AI-supported tools also allows for greater (digital) standardisation of engineering approaches. Instead of the current situation where asking 10 different engineers the same question would deliver 10 different answers, with agreed models, the same optimal solutions would begin to emerge. At this point, collaborative working brings even greater efficiencies, with all parties able to work and update from the same central source knowing that version control and accuracy is no longer an issue.

This approach is known as parametric modelling, leveraging AI and machine learning to find optimised solutions based on the various constraints mentioned above.

Smarter modelling

To illustrate the efficiencies that AI-based parametric modelling could deliver, imagine a project in which the initial plans called for a tank that could hold 100 cubic metres of water. The team would develop a solution which would feature retaining walls of the right thickness to ensure structural integrity while also satisfying all other key parameters.

Now, imagine a key objective or requirement changed, and the volume the tank needed to hold doubled. Traditionally, this would  mean starting again from scratch.

However, with a smart 3D model using parametric design, the new volume could simply be entered into the starting conditions and the tool would re-draw the design, with the appropriate wall thicknesses amended to account for the increased size and capacity.

The ability to make amendments in this way and have the model automatically re-configure the rest of the design to accommodate the changes generates significant time – and cost – savings for project teams.

It can also be used to visualise environmental impacts of the design such as carbon. The most cost-effective design solution may not be the most carbon efficient. Parametric modelling can be used to investigate various design solutions based on whole-life carbon parameters. This would allow the right information to be available at key decision points on the project and a little more investment on a better carbon reduced design solution could bring large reductions in the assets carbon footprint and cost saving benefits to the asset when it is in use.

3D modelling as a first step

In our experience, the UK water industry has another challenge that must be tackled before it can realise the potential benefits of parametric design: addressing the dominance of 2D modelling.

We estimate that 80 per cent of designs produced for UK water projects are 2D computer-aided design (CAD) drawings, with only 20 per cent using 3D models.

This naturally inhibits collaboration, with drawings sent out by a designer to all other parties for them to upload comments and share back with the original author for amends to be made.

Working in 3D right from the word go – as has become standard practice for our designers working in Nordic countries– allows more people to contribute within the same space, avoids clashes, and results in a more refined design with less chance of errors. 3D also helps all parties to understand any problems in a controlled environment before the team moves to the construction phase, where models can be used live to help guide the construction process. In the UK, many more issues with drawings remain un-noticed until this late stage, which can often cause headaches and delays.

Towards data-based design

The possible applications of this approach are endless, and include every aspect of water infrastructure design, from the structural elements that support it to flow modelling and energy management systems.

Ultimately, the goal is to move towards more design decisions being based on facts rather than intuition. AI is the perfect tool to allow engineers to achieve this.

The technology that enables this approach is still in its infancy, but it is developing quickly. Placing AI-based tools in the hands of designers and engineers will dramatically increase their ability to truly optimise projects, potentially revolutionising the way schemes are designed, reducing rework, digitally optimising design programmes, and making major savings.

The water and wastewater industries are under more pressure than ever to improve environmental performance and deliver maximum value for consumers, but within increasingly tight budgets. AI-based design automation could form a major part of the solution.

Smarter modelling is already helping to deliver optimised water and wastewater projects much more efficiently, something that will allow water companies to achieve improved outcomes for these critical schemes without breaking the bank. To find out how Sweco could help you reap the benefits of AI and wider digital thinking, contact Craig Hardingham today.