The JUNON programme aims to provide the Centre-Val de Loire region with a digital twin, which will enable the development of digital services to improve environmental monitoring and understanding, as well as the management of natural resources.
© JUNON
Should agricultural practices be adapted to cope with increasingly uncertain water supplies? Can we make regions more resilient to the risks of flooding or air pollution? How can we anticipate the effects of spatial planning on soil, water and biodiversity? The answers to these questions are currently based on fragmentary data and complex trade-offs between local stakeholders. However, a new approach is emerging: the use of environmental digital twins, which make it possible to numerically and automatically model the physical world over long periods of time.
In the wider environment (urban, agricultural, natural and industrial) and against a backdrop of climate change, where the management of natural and energy resources has become critical, these approaches are attracting growing interest from public and industrial stakeholders.
Making long term decisions on a day-to-day basis
Once deployed across a region, digital twins make it possible to adapt flows and practices in line with resource availability and risks.
For example, they make it easier to detect certain faults. These include detecting leaks in water networks, managing problem areas such as heat islands, and monitoring local pollution, particularly in relation to air quality. Conversely, they can also reveal systems with positive impacts, such as wetlands, which play a key role in maintaining biodiversity and promote water filtration and infiltration into the ground.
Digital twins can also be used to test different scenarios and plan future developments in response to various issues. Urban development, landscaping and environmental restoration projects can thus be analysed to assess their respective impacts and help local authorities make informed decisions. By providing an objective view of these effects, digital twins help to inform decision-making. They can also lower the temperature of often sensitive debates between local stakeholders.
Tools that are still difficult to get to grips with
The development of digital twins relies largely on recent advances in artificial intelligence (AI). Machine learning techniques now make it possible to continuously monitor and analyse large amounts of data and derive unprecedented predictive capabilities from it.
However, for many stakeholders (technical staff, elected representatives and the general public), this technological boost represents a paradigm shift that remains difficult to grasp. Can we rely on a model to guide practical decisions? How can we understand the assumptions and calculations behind the simulations? A lack of shared understanding can lead to differing perspectives amongst stakeholders and undermine twinning projects, which are often lengthy and costly.
Support is thus essential. Feedback on other regional digital twins, training courses and outreach initiatives are helping to facilitate the gradual adoption of these tools.
Data-hungry
Digital twins depend on the data they have access to. To represent and monitor complex and extensive environments in real time, they require vast amounts of information and measuring equipment that is often expensive. This is the case, for example, with certain water quality monitoring-probes, which are expensive to buy and require regular maintenance.
Some variables are difficult to measure automatically, such as biodiversity, which still relies heavily on field observations. The frequency and scope of the surveys are limited because they must be carried out by specialists capable of undertaking the lengthy and laborious task of counting the various species present. Similarly, certain emerging pollutants, such as PFASs or some pesticides, still have to be measured in laboratories.
To overcome these limitations, digital twins often rely on “proxies”. These are variables that are easier to measure and are used as indirect indicators of a phenomenon because they show a strong correlation with the target variable.
Twins which are sometimes repetitive and poorly coordinated
Although these technologies are still being developed at a regional level, numerous projects are already emerging, spearheaded by a variety of stakeholders: local authorities, public agencies, businesses, engineering consultancies, and so on.
However, these initiatives are often developed independently. It is not uncommon for several digital twins to utilise similar data streams or offer similar functionalities, without any real sharing of resources. One of the key challenges is therefore to make these systems interoperable, which is to say, capable of communicating with one another and sharing data and results.
Digital twins are now becoming an increasingly ubiquitous tool for regional management. They offer innovative solutions, but present numerous challenges that will need to be tackled collectively.
