European collaboration to help design Smart Water Futures

A new European collaboration, backed by an ERC €10 million grant, is on a mission to find out how we can design urban drinking water systems of the future. We speak to the four prominent European scientists leading the Water-Futures initiative to find out more.

Designing next generation urban water systems

How will the world provide high-quality drinking water services in a future facing increasing climate change, economic and population pressures?

How do we design and create the next generation of intelligent urban drinking water systems?

A new cross-European research project is on a mission to find out the answers to these two pressing questions and more.

Funded through the European Research Council's Synergy Grant, a total of €10 million will be provided from mid-2021 over a period of six years.

Entitled ‘Smart Water Futures: Designing the Next Generation of Urban Drinking Water Systems’, the project will be led by a Water-Futures team that combines leading experience on water science, systems and control theory, economics and decision science, and machine learning.

Four prominent European scientists will lead the Water-Futures initiative, including:

Professor Dragan Savić, CEO of KWR Water Research Institute (KWR), the Netherlands and Professor of Hydroinformatics at the University of Exeter, UK
Professor Barbara Hammer, Machine Learning Group at Bielefeld University, Germany
Professor Phoebe Koundouri, Professor of Economics at the Athens University of Economics and Business, Greece
Professor Marios Polycarpou, Director of the KIOS Research and Innovation Center of Excellence and Professor of Electrical and Computer Engineering at the University of Cyprus.

This project aims to develop a new theoretical framework for the allocation and development decisions on drinking water infrastructure systems. One of the target outcomes includes an integrated decision and control framework to be implemented as an open-source research toolbox.

New science outcomes will be applied to three case studies exemplifying different types of urban water systems:

A mature, relatively stable system
A mature and rapidly expanding system
A relatively recent supply system in a developing country with high growth and special challenges, including limited resources, intermittent supply and high water losses.

A holistic decision-making framework

The project aims to combine long-term water thinking to cope with an unpredictable environment with the short-term decisions needed to achieve smarter, more efficient operations.

“The challenge of designing future-proof drinking water systems is one of those so-called “wicked problems”, which are difficult or impossible to solve because of incomplete, contradictory, and changing requirements that are often difficult to recognise,” said Professor Dragan Savić.

He cited the cases of booming and shrinking populations as one example of uncertainty when long-term planning for water demand over 100 years.

For example, Shenzhen in China rose from a fishing town of 30,000 people into a megacity of 13 million inhabitants over 40 years. Meanwhile, in 2019 Japan’s natural population declined by 500,000 people.

“The challenge of designing future-proof water systems is one those of so-called “wicked problems”, which are difficult or impossible to solve.”

“That is difficult to take into account when planning long-term investment in city water infrastructure even without the uncertainty brought about by the prospect of climate change,” he added.

The CEO believes there is a need for a more holistic and intelligent decision-making framework for managing water infrastructures in the cities of the future.

"Sustainable planning and management of water infrastructure are definitely one of those problems. For example, drinking water infrastructure is supposed to last for over 100 years, but we do not know how the city they serve will develop over time and, consequently, how the demand will change over those 100 years,” he added.

A human-centric approach

As well as the four research organisations, Water Futures will engage relevant stakeholders globally through living labs.

For implementation, it will use a systems innovation approach to co-design future urban water systems by co-developing the necessary technological, policy and financial pathways.

Professor Phoebe Koundouri said there remain many unresolved scientific questions regarding the “human-centric approach”.

As a result, researchers will drive for “an alternative, deeper, more mature understanding of the structure of human preferences and decision-making process, over private and public goods, the short-run and the long-run, and under risk and deep uncertainty”.

“We need a unifying framework for short and long-term decisions.”

This will include "mathematical modelling and (virtual reality) experimental validation, of people's choices regarding public goods, such as infrastructure and related economic-social-environmental costs and benefits. This process will produce new insight for the provision and pricing of water services".

Koundouri said: “We need a unifying framework for short and long-term decisions. This must integrate the correlation between time and uncertainty and be derived from a deeper understanding of the structure of human preferences.”

Three breakthroughs for machine learning

From a machine learning perspective, the project will enable three breakthroughs:

To bring forward research on what actually 'explainability' means in machine learning
To progress methods for complex, non-vectorial networked data and machine learning will blend with technologies as offered by control theory and evolutionary strategies
Demonstrate the potential and challenges of machine learning methods in a complex problem for uttermost practical relevance by looking at the short and long-term water infrastructure.

“We will face problems that will become moving targets due to changes in the environments and demand. We need to devise robust methods to deal with these uncertainties under continuous changes,” said Professor Barbara Hammer, Machine Learning Group at Bielefeld University.

“Furthermore, we will need to deal with an extraordinarily complex problem, where intelligent methods need to be designed to tackle computational complexity and enormously large temporal dimension and wide spatial range.”

Incorporating information and communication technologies

From a technology viewpoint, the proliferation of sensors and actuators of various types, in the framework of the Internet-of-Things (IoT), in conjunction with the roll-out of 5G wireless networks, will facilitate the acquisition of far more data related to water systems.

Professor Marios Polycarpou predicts that in the near future, we will be seeing the wide integration of information and communication technologies with water systems, forming a large-scale and complex “cyber-physical system”.

“One of the key objectives of the project is to utilize these new computing, sensing and communication technologies to develop intelligent automation methods for monitoring and control, that are evolvable as sensors and actuators are added or removed.”

“Ultimately, the project will contribute to the design and operation of the next generation of smart water systems.”

Professor Polycarpou said: “Ultimately, the project will contribute to the design and operation of the next generation of smart water systems, which will integrate real-time monitoring and control with long-term robustness and flexibility, while incorporating economic, social, ethical and environmental considerations.”