On Thursday I had the pleasure of hosting a panel on AI & Climate at TechUK's NetZero conference. It was a thought provoking event with leaders, thinkers and industry enthusiasts attending some important and interesting talks. It was actually one of the most interactive of conferences I've attended in a while, with large amounts of audience participation and a general buzz of energy around what we can achieve in our race to Net Zero.
AI’s role in addressing climate is enormous and we discussed how AI has been used, and could further be used, to positively impact our drive to net zero and decarbonisation. I was joined by three incredible panellists who are all extremely experienced in the space and great thinkers alike. We had Simon Bennett, Director of Innovation at AVEVA, Tarja Strahlman, Energy Transition Practice at Faculty AI, and Henry Franks, CTO of Climate Policy Radar.
It’s been acknowledged that AI is a game changer for the net zero transition. But for all the hype, the advice from our panellists was to focus on the problem at hand and apply AI as a solution.
During our session, we discussed how AI can facilitate the reality of a smart grid to provide and receive energy. Tarja talked about her experiences of the grid transition and how AI is the enabler in making this a reality. A very real example of this is the curtailment of wind energy alongside the need to manage the load through storage vs delivery optimisations.
Simon talked through his experiences using AI to expedite process simulation in the manufacturing space. This is where AI can be used to rapidly simulate real world scenarios to both increase speed but also quality, through the exhaustiveness of decision making consideration and calculation. We also discussed how this can manifest into a digital twin and even the visualisation/metaverse, to expedite virtual world gamification of collaboration. The real-world impact of this is enormous, reducing the material and energy consumption required to find the right solution.
We discussed plenty of other use cases: using AI to monitor energy distribution to ensure consistent supply and lessen wastage, reducing the cost of maintaining critical infrastructure, and predictive capabilities to reduce the number of environmentally damaging incidents.
The ability of AI to augment the capabilities of humans and make their lives easier shouldn’t be underestimated. Pointing it at low value tasks increases the time for high value tasks, making people more strategic and thereby giving them more value in their careers. Henry’s example was a poignant one: imagine a worker scanning thousands of regulatory documents vs training an algorithm to perform scalable search which is fine tuned by humans. He rightly called AI a "Force Multiplier".
From our experience, the augment concept is true across every large energy & utilities company. Businesses that have been built around human decision making are naturally ripe with data availability, and there is a huge opportunity to use AI to augment how these subject matter experts perform their daily decision making with the support of AI.
Creating insights from available data is nothing new and an approach we're working on across multiple industries, whether that's for highly regulated financial services institutions or making data more strategic asset for energy major TotalEnergies.
Through a number of the day’s sessions, there was a clear consensus on the impact of data to transition to net zero. There is an awareness around the challenge of having highly available, connected and trusted data. Without this, the opportunity to utilise AI is limited.
This is acknowledged as the single biggest challenge to AI adoption, in line with the findings of our State of AI in the Enterprise report. While it requires considerable effort to make data accessible and available, building data platforms for AI to run on is the answer. For energy & utilities firms, with enormous and highly valuable data estates, making data accessible can build trust and integrity and enhance how AI can be applied across the enterprise.
At Mesh-AI, we have long said that AI is only worth the data it's built on, which is why our data strategy accelerator helps enterprises to reimagine their data estate and improve the quality of their data.
The other big barrier to AI adoption at scale is trust. In a world filled with hype, there is a large level of distrust around the use of AI, especially in the areas of bias and transparency.
We discussed how the answer to this is largely centred around literacy and education. The buzz around GenAI has driven a huge awareness increase but led to a narrowed view on what AI is, dismissing most traditional AI that came before. This distortion in the perception of what AI is and what it is capable of, can make it harder for enterprises to adopt AI at scale and harness the benefits. Focussing on outlining the way AI works, the options available and the solutions to risk, will increase trust and adoption across the sector.
Our research report found energy firms are lacking in AI maturity–30% of energy firms are pessimistic about their AI maturity compared to just 9% of financial services organisations. Additionally, 21% of energy enterprises are not confident they have the right skills to safely and securely innovate with AI.
So while the energy & utilities industry may well be lagging behind other industries, the energy and appetite at TechUK’s Net Zero conference demonstrates the opportunity for AI to power the clean energy transition is enormous. We discussed a number of recommendations for enterprises on their road to AI adoption:
Measure: You can't improve what you can't measure. AI as a solution is optimised against a set of metrics, so in order for us to point AI in the right places, we need to design the right metrics and build them into any system or process.
Speed: Identify the areas of opportunity for increasing the efficiency of processes. These are often non-obvious AI applications, and can begin with simple Robotic Process Automation, but can also lead to AI-based capabilities to improve these tasks: e.g. prioritisation, sequencing, execution.
Capability: Apply AI to tasks that humans cannot do or would take years to do. For example, AI can consider all possible scenarios of an outage across the grid in seconds, considering a multitude of factors all at once, such as the demand, component age, and resource in question. These AI-powered insights can inform human decision making around energy usage, emissions, infrastructure health and grid performance, and enable the clean energy transition.
Find out how Mesh-AI revolutionised how this energy firm approaches data, powering its net zero transition.
AI and digital twins have a huge role to play in the race to net zero.
Read our key takeaways from AWS' latest Energy Symposium.