please enter please select

NEWS & INSIGHT | Opinion

Path to net zero: AI can hold our hand

02 December 2021

Achieving net zero by 2050 is essential to preserve our way of life. Artificial intelligence will have a fundamental role in achieving our net zero future – by providing the pathway and decision making required. We have a moral obligation to ensure that the transition from hydrocarbons to renewable energy sources doesn’t just happen but is done smoothly without major economic and societal disruption. We will need to make many informed decisions, during this transition and artificial intelligence will help us deliver a smooth path to net zero.

Recognition of talent

As part of my role at the Net Zero Technology Centre, I have the privilege of seeing truly innovative and ground-breaking technologies and platforms that seek to deliver an affordable net zero energy industry. In January 2022 we are proud to be supporting the RangL challenge. The rangl competition platform, created at The Alan Turing Institute to serve as a new model of collaboration between academia and industry, invites participants from the global AI community to create and apply their own reinforcement learning (RL) algorithms to solve a ‘Optimal Pathway to Net Zero by 2050’ challenge created by associates of The Alan Turing Institute, Oxquant and the Offshore Renewable Energy Catapult.

This type of approach is known as crowdsourcing, and it is not new. Many multi-national companies have used this approach of inviting the global community to solve complex problems, and in many cases, it has delivered excellent results. Modern-day advances in computing have significantly enabled this approach – particularly cloud computing – democratising the playing field and allowing users without access to powerful computers the opportunity to take part in these types of challenges but importantly providing peer recognition for their talent. The RangL challenge aims to bring together the AI community to raise awareness of the biggest challenge of our lifetime and demonstrate the power of crowdsourcing.

Realise the path

So, what exactly is the challenge? To determine this, the first step is to understand the environment in which the challenge exists. Using inputs from the Reimagining a Net Zero North Sea: An Integrated Energy Vision for 2050 report, the team created a set of rules and assumptions for the environment in which the RL agent will exist. As a foundation, the environment will use the three outcomes highlighted in the report (emerging, transformational and progressive) as baselines from which a range of scenarios can be determined by an RL agent created by the participant. An RL agent is fundamentally a program that learns in much the same way humans and animals learn – through cause and effect. It performs an indeterminate number of actions from which it is rewarded (or reinforced), a process of fine-tuning to deliver the most rewarding outcome.

The overall objective of the challenge is to find optimal scenarios in upscaling offshore wind, hydrogen (blue and green), and carbon capture and storage to manage the transition from traditional oil and gas and inform the decision-making process regarding CAPEX costs, OPEX costs, decommissioning costs, jobs, revenue, and emissions. Fundamental answers to whether we can accelerate the upscaling of these technologies, what are the major obstacles, do we have capacity and what will it mean for employment will be sought. The challenge may not answer all the questions, but there is a need to fully harness the benefits of crowdsourcing solutions from a global AI community – a view shared by a recent publication by Microsoft and PWC, How AI can enable a Sustainable Future, which stated that “all stakeholders across the public, private and third sectors must be involved in unlocking AI to tackle environmental challenges to its fullest potential.”

In the real world, the optimal pathway will likely not be achieved, but we can certainly gain a picture of what that could look like and how it may be achieved to come close to achieving that pathway. We need to use every available tool to inform the journey, AI can hold our hand along the way.

Find out more about the RangL challenge here.

Subscribe for the latest updates