Recently, digital technology has been the subject of controversies over its carbon footprint. While it is not free of reproach, the digital sector is also a powerful factor in decarbonisation: as we are dangerously close to the tipping point - the moment when the impact of global warming will be such that its consequences on life on Earth will be irreversible - it is urgent that the digital sector, like the others, take its part in the transition.
One doesn’t necessarily perceive Machine Learning technologies as a driving force to decarbonize the human activities. Nonetheless, it is interesting to notice that the environment paradigm is generally both multifactorial and sequential. Tackling environmental externalities requests to address numerous factors, simultaneous and sequential. For instance, decarbonizing a supply-chain in the automotive industry implies to work on many parameters : quality of supplier, transparency, speed of delivery, default risks, stocks, etc. Each of these factors being interdependent with the others. Interestingly this complexity is inherently what Machine Learning is good at : addressing complexity, multifactorial and sequential at the same time.
Laitao helps organizations to analyse in which dimensions of their businesses they can use Artificial Intelligence and Machine Learning to increase their productivity while decreasing their environmental externalities.
GREEN IT
Current debates place the share of CO2 emissions related to digital technology at between 2 and 4% of total emissions, with particular concern about the annual growth in digital consumption: if nothing is done, this could quickly exceed 6%. There is already a large body of literature on what is now called Green IT, i.e. how to rapidly reduce the pollution of terminals, data centres and networks. One of the avenues is the fight against device obsolescence and the optimisation of recycling.
Different estimates of the digital share of CO2 emissions
IT FOR GREEN
A less talked-about subject concerns IT for Green, which is at the heart of Gilles Babinet’s work, notably at the Institut Montaigne.
Digital technology allows performance gains. These improvements are not negligible and show that beyond studying the issue of green IT - energy consumption, CO2 emissions and other social and ecological externalities of digital technology, it is important to see how digital technology can be used to serve the environment in an IT for Green approach. Through numerous examples (smart electricity grids, waste treatment and recycling, supply chains, industry 4.0, housing and urban planning, transport, etc.), the significant performance gains made possible by digital technology are highlighted
However, for digital technology to become a real factor of transition, these developments call for cultural, regulatory and technological changes. Digital technology must get out of the consumerist logic in which it has become locked up! There is an urgent need to make the population aware of the urgency of the situation. This can be done, for example, through the systematisation of impact measurements which will eventually make it possible to know the true price of CO2. It is also important to increase education on the subject because there is a terrible lack of skills in this area. The “datacologist engineer”, specialist in machine learning applied to the environment, could well be the skill that tomorrow everyone will be looking for.
EDUCATION AND INDUSTRY
This necessary transition is a real opportunity for France, especially for the country's industry. On a national scale, the opportunity is expressed in hundreds of thousands of jobs, provided that France chooses its fields of action carefully.
However, to think for a moment that we are going to make a total energy transition without profoundly changing our uses is naive at best. The stakes are so high that it will require a major reorganisation of our human societies and our uses. This post only evokes one of the fields of action to achieve this transition.
The issues related to this challenge of combining digital and green transition are at the heart of Laitao's work.
Gilles Babinet often explains the role of digital technology in the ecological transition through conferences. Beyond studying the issue of green IT - energy consumption and CO2 emissions and other social and ecological externalities of digital technology, it is important to see how digital technology can serve ecology in an IT4Green approach. Through numerous examples (smart electricity grids, waste treatment and recycling, supply chains, industry 4.0, housing and urban planning, transport, etc.), the significant performance gains made possible by digital technology are highlighted. It also discusses the cultural, regulatory and technological changes that need to be made for digital to become a real factor of transition.