Water management should first create a big data base

The question"Can machines think? ’ was asked by Alan Turing in 1950. With this question, Turing opened the discussion on machines with human-like intelligence that can learn. In the 1970s, due to the increase in computing capacity, the problem-solving ability of computers progressed a lot. However, this progress was still far from imitating human intelligence and, more importantly, the way it learns. By the 2000s, big data and increased processing power accelerated artificial intelligence (AI) studies. The development of AI capabilities accelerated with the significant advancement of artificial neural networks in 2012. These networks enabled machines to learn by reinforcement and replicate the way the human brain processes information. With techniques such as machine learning and deep learning, AI systems have emerged with human-level performance models such as image recognition, natural language processing and chess. A summary of the historical development of AI is given in Figure 1.

Artificial Intelligence in Water Management

The world's population has increased 4 times compared to 100 years ago and water authorities are struggling to provide drinking and potable water in a sustainable manner, especially for the urban population.

In addition, water plants and wastewater treatment plants are operating with outdated systems and have difficulty in responding to increasing needs. The business approach that depends only on human labour force does not allow increasing efficiency and reducing water costs in these enterprises due to the lack of monitoring systems and tools to make analyses.

The use of AI in water management has great potential both to ensure efficient use of resources and to prevent water crises. For this reason, the use of artificial intelligence in water management should be spread consciously. While preparing the technical and technological infrastructure for this, it is important to train personnel trained in the recognition of artificial intelligence algorithms and the analysis of their results. Of course, the need for healthy, continuous, standardised data arises for progress in this field. Since this data collection process is time-bound, it should be started as soon as possible. In addition, the available data should be analysed and processed in a way that can be used.

In addition to the benefits obtained and planned to be obtained from artificial intelligence, issues such as the development of reliable and responsible AI systems and the interpretation of the results obtained from artificial intelligence in the most beneficial way are the prominent issues in the artificial intelligence sector.

It has been observed that the algorithms used in automation systems are open to many biases. Therefore, reliable, explainable and transparent practices are vital in the use of AI systems. Turkey is among the leading countries that have published an AI strategy and is implementing actions in coordination with this strategy.

The Importance of Data in Artificial Intelligence

The success of AI systems largely depends on the quality of the data used. It is of great importance that the data used are obtained from accurate, up-to-date and problem-solving datasets. AI applications may produce incorrect results due to deficiencies in training data or biases and biases.

As a precaution, data sources should be carefully selected, continuously updated and data cleaning processes should be carried out meticulously. Therefore, in order to obtain accurate and useful results in artificial intelligence, it will be necessary to pay attention to issues such as creating a healthy, reliable, continuous database and selecting the right datasets.

In addition, since AI systems consist of systems that continuously learn and adapt to environmental changes, the system may develop in unexpected ways and control may become difficult. As a precautionary measure, the AI life cycle, including data, models and outputs, needs to be continuously monitored and maintained more frequently in order to prevent the system from exhibiting different behaviours than expected.

Artificial Intelligence in Water Management

AI, especially with its rapid rise in recent years, deeply affects almost all sectors. Thanks to advanced algorithms and big data analyses, it has started to be used in a wide range of areas, from increasing the efficiency of businesses to optimising production processes. Water management is one of the areas where artificial intelligence can be used effectively.

The benefits of AI in water management can be summarised as follows:

With the use of artificial intelligence in water management; benefits such as development of effective management and early warning systems with more accurate data, better infrastructure planning and prevention of water supply problems, reduction of water losses and cost savings, cleaner water resources and sustainable water management, reduction of the effects of crises and rapid response and increasing water saving and conscious consumption behaviours of individuals can be achieved.

Monitoring:

AI monitors water quality, flow, temperature and other parameters in real time by analysing data from sensors and IoT devices. AI-based models can predict water movements, groundwater levels and recharge processes in basins.

Planning and Demand forecasting

AI forecasts rainfall amounts, drought or flood risks by analysing meteorological data. It estimates water demand based on variables such as past usage data and population growth.

Loss and Leakage Detection and Optimisation of Water Distribution Systems

AI uses anomaly detection algorithms to identify leaks and leakages in water networks. Water pumps and treatment plants are optimised for energy efficiency.

Efficient Treatment and Reuse

AI optimises chemical dosing, filtration and other treatment processes. Management of processes for the reuse of treated wastewater is provided.

Decision Support Systems for Risk and Crisis Management

Artificial intelligence can develop rapid response systems for natural disasters such as floods and droughts. AI suggests managers to make the most effective decisions by simulating different scenarios.

Artificial Intelligence Applications

In order to ensure positive development outcomes from AI applications, capacity and infrastructure development policies for AI utilisation should be addressed holistically.

Capacity building policies in this area should fulfil the basic needs of all water-related stakeholders in terms of AI-related skills development in artificial intelligence and Information and Communication Technology (ICT) needs in energy, data generation and storage.

Policies in the water sector should implement training and adaptation efforts to provide a qualified workforce to accompany AI-led innovations in water, while also implementing policies to reduce the wastage of the existing workforce by increasing investments.

Artificial Intelligence Security Threats

There are great expectations about Artificial Intelligence and Machine Learning. However, there are ongoing debates about where this path will take us. For this reason, I think that security and reliability are issues that should remain at the forefront of the discussion, because a greater dependence on digital tools and automation will also magnify and increase the negative consequences of a security breach or system failure.

For example, in addition to the benefits obtained and planned to be obtained from artificial intelligence, there are also negative situations that reveal the need to develop ‘reliable and responsible AI systems’. In some applications, it has been observed that the algorithms used in automation systems are open to many biases. Therefore, reliable, explainable and transparent practices are vital in the use of AI systems.

The EU has published the AI Law in order to prevent risky situations caused by AI and put it into effect as of 1 August 2024. It aims to protect EU citizens from potential AI dangers by ensuring that AI systems used within EU borders remain within the framework specified in the law. Global companies, for example smartphone manufacturers, preferred not to activate some of the AI features in their new models within the EU borders with the argument of ‘regulatory uncertainties’.

In order to co-ordinate the work on AI technologies in our country, the NIRZS was published. In order to achieve the targets set in the strategy, the Action Plan was updated and renewed to cover the years 2024-2025. When the Action Plan is analysed, the activities to be carried out on the public side and the anticipated benefits to be obtained are seen. Actions also include risk-based management of AI projects and certification with the ‘Trusted AI Stamp’.

Conclusion

We know that technology cannot solve all our problems. However, we still need digital solutions offered to us by technological developments. For example, with rapidly developing quantum computers, machine learning and artificial intelligence elements will gain a much smarter and faster working quality as a result of analysing a lot of data in a short time.

Technological developments reveal that the digital journey of water will gain momentum. Technological advances can help solve water problems for the benefit of all humanity and create a more sustainable world. We have seen that some metropolitan municipalities in our country have taken initial steps for this digital journey. It is important to continue these efforts with an understanding that prioritises social interests and our national interests. We also need to prepare for the management of this rapid change in every field. The primary steps are the development of domestic software in this field and the preparation of digital transformation technical infrastructure and human resources in our institutions for this process.

Artificial intelligence has the capacity to provide more efficient, environmentally friendly and sustainable solutions in water management. However, infrastructure investments, data sharing, interdisciplinary cooperation and training programmes are required for the effective use of technology. Therefore, prior to the use of artificial intelligence, it is necessary to prioritise personnel training as well as technical infrastructure and large database creation.

Studies on the most effective utilisation of artificial intelligence by making preliminary preparations in Water Management will make a great contribution to the protection of water resources and the management of water crises.

References

[1] Report on Artificial Intelligence Applications in Public Sector. Informatics Association of Turkey November 2024

[2] Yıldız D. Özgüler H. 2020. Digital Water World Report. Water Policy Association. Report No:29 .19 May 2020

[3] Yıldız D. Özgüler H. 2020 Artificial Intelligence and Water Management Report. Water Policy Association. Report No:30 4. July 2020