Smart City at your service - empowered by emerging technologies Vol. 3
Publication Date/Time
2018-09-12T09:00:00+00:00
How Open Data and Artificial Intelligence help to manage water
resources better in smart cities
OPEN DATA AND ARTIFICIAL INTELLIGENCE - A GOOD INTERACTION

Water is one of three basic needs for humans. Our earth is mostly
covered with water, while only two percent is fresh water. However,
this topic does not only concern cities, but intelligent water
management is one of the most vital challenges for smart cities in the
future. With Open Data, it is possible to compare the water resources
of countries as well as cities and derive measures from that.
Regarding resource water management, the combination of Open Data and
Artificial Intelligence (AI) builds a good interaction to help
managing it. Read more about the symbiosis of Open Data and AI, in the
article How Open Data and Artificial Intelligence improve the air
quality in smart cities.

NOWADAYS, THE WATER SHORTAGE IS ALREADY FELT

For a very long time, the sedentary nature of humans has been
determined by water. This has mostly changed in developed countries
but not so much in developing ones. Around the globe, forty percent of
people are affected by the changes in the climate, including the
reduction of fresh water. However, water shortage does not only appear
in Africa but also recently within Europe. Since this summer provides
hot temperatures for a long period of time, the German press
[https://www.stern.de/panorama/wetter/trinkwasserknappheit-in-deutschland--es-droht-die-duerre-8201534.html]
announced that in certain parts of Germany a drinking water shortage
my come soon if the weather remains the same.

IN THE FUTURE, THE DEMAND FOR WATER WILL INCREASE 

Scientists expect a temperature increase on the earth of five degrees
Celsius. This, in connection with desert spread, river dehydrations
and groundwater pollution leads to an increased importance of water.
By 2025, the UN Food and Agriculture Organization
[http://unesdoc.unesco.org/images/0026/002614/261424e.pdf] states that
drinking water will be insufficient for all people on earth. One cause
is the population increase - by then the population will grow by 2.5
billion, all to be provided with drinkable water. Indeed, this topic
does not only apply to remote countries but will also affect Europe in
the future. By 2070, forty-four million Europeans
[https://www.wiwo.de/technologie/umwelt/weltwasserforum-wasserknappheit-in-europa/6318760-2.html]
will be affected by water shortage because rivers in Central and South
Europe will carry eighty percent less water - even power plants may be
on dry land. Therefore, the UN
[https://www.unric.org/en/latest-un-buzz/29171-more-valuable-than-oil-diamonds-or-gold]
states that in fifty years, drinking water will be more valuable than
gold.

THE CONSTRUCTION OF DAMS, DEEP WELLS OR WATER PIPES WILL NOT SOLVE THE
PROBLEM, BUT NEW TECHNOLOGY WILL

The urban water supply industry
[https://www.waterworld.com/articles/iww/print/volume-16/issue-4/features/artificial-intelligence.html]
should be guided into the new technological era. Predominantly, the
current infrastructure of the water industry still needs a human
engineer and/or operator for anticipating and understanding how to
best react to recurring changes in inlet water quality, e.g. episodic,
daily or seasonal, hoping for achievements regarding sustainable
operations. Nevertheless, also a well-trained engineer and/or
operator, who is familiar with big data, needs time to analyse and
predict, which system changes need to be done. A solution for this is
the combination of new technologies such as Artificial Intelligence
(AI) and Open Data. The idea is that urban water suppliers include
data-driven technology
[https://waterfm.com/artificial-intelligence-reshaping-water-sector/]
allowing them to find previously unattained information regarding
water demand and supply.

There are many ways of how AI replicates the human learning process.
In the AI learning phase, input data is coincided to well-known
outputs allowing the algorithm to learn. As a next step, during the AI
operational phase, the system starts with the introduction of new
data. Since AI is permanently able to process a huge amount of data in
real time, it is the perfect instrument for water resource management
in a constantly changing world and the water business. This allows
water suppliers to achieve higher revenues as well as plan effectively
for the next years.

AI CAN SUPPORT (SMART) CITIES TO MANAGE THEIR WATER RESOURCES MORE
EFFICIENTLY IN MULTIPLE WAYS 

As the figure
[http://www3.weforum.org/docs/Harnessing_Artificial_Intelligence_for_the_Earth_report_2018.pdf]
below displays, AI can improve the water management within (smart)
cities regarding water supply, catchment control, water efficiency,
adequate sanitation and drought planning.
[https://data.europa.eu/sites/default/files/img/media/20180912-vol3-water.png]
RESOURCE VOLATILITY, SUPPLY OPTIMISATION AND AUTOMATED SUPERVISORY
CONTROL ARE EXAMPLES OF HOW AI CAN HELP TO MANAGE WATER RESOURCES
BETTER

Water suppliers have two options: either search for new supply, which
deals with the water resource volatility, or optimise the existing
supply. Rome, well-known for its historic aqueducts, had to ration
their fresh water
[https://www.washingtonpost.com/gdpr-consent/?destination=%2fnews%2fworldviews%2fwp%2f2017%2f07%2f28%2fromans-are-about-to-go-eight-hours-a-day-without-water%2f%3futm_term%3d.7be9ddb96de3&utm_term=.b404456acbdb]
this summer due to the drought. One reason for that is their leaking
urban water infrastructure, which loses about forty-four percent of
the fresh water in their pipes. AI can help to solve this issue in
applying predictive analytics to the legacy water equipment as well as
data management systems. The data from disparate sources, e.g. pumps
or filtration systems, run through algorithms and delivers optimal
control and management protocols, which then can be used to conserve
energy.

In using new Software-as-a-Service
[http://dro.dur.ac.uk/624/1/624.pdf] platforms and AI, better
financial, strategic and tactical utility operations can be created.
With AI
[https://waterfm.com/artificial-intelligence-reshaping-water-sector/],
municipalities can gain a better understanding of real-time water loss
and a more efficient distribution network - while coping with the
rising customer demand in their cities.

Another area, in which AI supports the water resource management is
concerning automated supervisory control. Via a system, data from
different sources is evaluated and based on that, the water pumping
from the aquifer is regulated. This data contains a mixture of
weather- and system-related data in order to optimise the fresh water
supply. According to waterworld.com
[https://www.waterworld.com/index.html] "pairing these automated
controls
[https://www.waterworld.com/articles/iww/print/volume-16/issue-4/features/artificial-intelligence.html]
with an extended knowledge base will lead to a network of systems that
can automatically communicate with each other and acclimate when one
system requires more resources than another in order to keep operating
costs as low as possible."

IN THE FUTURE, AI WILL SUPPORT WATER RESOURCE MANAGEMENT EVEN MORE

Soon, urban water supplier will use AI to identify and quantify
various types of cyanobacteria
[https://phys.org/news/2018-07-artificial-intelligence-technology.html#jCp],
or blue-green algae, threatening to shut off water systems when they
proliferate. AI combines a software and a microscope to analyse water
samples for algae cells within one or two hours. This includes already
that a human analyst confirms the AI results. Currently, samples are
sent to labs and manually analysed by a technician, which takes one or
two days. Since water sample testing would be quicker and can be
conducted more frequently, AI would offer an early warning system of
occurring issues. The aim is to provide an AI-based system, which
permanently monitors the water flowing through microscopes to track
contaminants as well as microorganisms.

Another futuristic approach
[http://www3.weforum.org/docs/Harnessing_Artificial_Intelligence_for_the_Earth_report_2018.pdf]
of how AI can help to manage water resources is based on machine and
deep learning. With the household smart meters, a huge data set is
generated, which may be used for water flow predictions, inconsistency
identification and/or leak controls. As a next step, machine learning,
Internet of Things and blockchain will be combined creating a
decentralised water system in which local water resources and closed
water loop recycling add value. Giving an example, rivers can be
designed "to autonomously adjust
[http://www3.weforum.org/docs/Harnessing_Artificial_Intelligence_for_the_Earth_report_2018.pdf]
their own sediment flow". Pairing this with AI-informed price models,
it may optimise the usage of water and change people's behaviour.

Therefore, in AI lies the future of the water industry, better to say,
water resource management. It may play the pivotal role to make our
present water infrastructure more robust and safer. However, AI can
only be as good as its underlying data is, therefore, Open Data plays
a significant part in this process.

Read more about how AI and Open Data can support smart cities in terms
of clean air
[/en/highlights/smart-city-your-service-empowered-emerging-technologies-vol-2]
on the European Data Portal.
