Skip to main content

Green skills challenge gets worse- will AI help?

 The renewables revolution demands up-skilling, with there being shortfalls in electronic, electrical, technical, & engineering and other key skills.  For example,  the recent UK Engineering Construction Industry Training Board’s (ECITB) Sectoral Workforce Census report notes that 81% of renewables employers in the ECI sector are experiencing challenges hiring workers, compared to 71% in the wider ECI in Great Britain. Electrical and mechanical fitters, pipefitters, platers, non-destructive testing technicians and planners are among the roles that are proving most difficult to recruit.  

That is worrying, since employment in conventional energy fields is declining as the use of fossil fuel is being phased out. In theory that means there ought to be plenty of takers for the new green jobs. The ECITB report highlights the unique opportunities in the six key renewables sub-sectors, with biomass being the largest construction sub-sectors, accounting for 24% of the renewables workforce, followed by energy from waste (21%), offshore wind (20%), biofuels (15%), onshore wind (14%) and solar (8%). But there are significant challenges in making the job conversions, including the provision of skill retaining and updating facilities.  

Indeed, the National Infrastructure Commission (NIC) says skills shortages ‘pose a significant challenge for the deliverability of the energy transition’ from fossil fuels to renewable energy. In particular it says the UK needs ‘a step change in the approach to the distribution network’, with Construction News reporting that the sector was already suffering from a shortage of craft workers, particularly cable jointers and overhead line workers, as well as engineers and transferrable construction roles such as excavators and project managers. NIC pointed out that ‘given the time it takes to develop these skills, urgent action must be taken to ready the workforce for the energy transition and to meet net zero. Short and medium-term solutions – such as retraining and recruiting skilled workers from other countries - can help, but they will not be sufficient to manage the long-term workforce challenge’.  

Note that few of these skills are likely to be suited to AI substitution- so AI may not been able to help much with skill gaps, at least not yet. In general, although NIC says that in all sectors ‘there are common challenges with finding sufficient engineers and other skilled roles, such as data and digital specialists’, it is likely that the situation for most metal bashing /hands-on engineering activities is very different from that for market research, accounting and management- or even design staff. The later may well find AI taking over, in time, unlike fitters, mechanics and ‘white van’ technical repair/maintenance people. 

However, it may all begin to change as AI develops. Although it depends on what we mean by AI. It is often used as a catch all term, to cover many different things. ‘Automation’ is maybe better used term for computer management or control of processes, and that can be important for some manufacturing, e.g. of solar cells and solar and some wind system components. So that is likely to be relevant now in the energy sector.  Some robot systems might also be able to handle some types of routine repair and maintenance work.  But for more complex jobs and more advanced skills, we may need more. 

A distinction is often made between two types of AI. Generative AI, used to create new content, such as ChatGPT, based on the use of large language models – but that’s not directly relevant to the energy sector. Unlike Predictive AI, which is used to forecast patterns and identify future trends based on existing data that can provide key benefits to the energy system.  Fortunately, that is less energy-using than Generative AI, as a new review by POST  makes clear. And it says that ‘AI is already being developed and used in the UK grid, predominantly by the national transmission network and energy system operators for forecasting and maintenance’. It adds that AI ‘has the potential to leverage big data from devices such as smart meters and other technologies to optimise energy planning, generation, storage and use’.

So, it may be that some types of AI can in time help relieve some skill shortages, and they can also offer better ways of doing things and possibly open up new skill areas enhancing rather than replacing human input. However, although AI can do many things well and fast, there are limits.  AI is expensive and although it varies with the type and the job, it can be very energy using- so we may have to use AI sparingly, and not for easy problems/activities that can be dealt with/managed with simper computers- don't use a sledge hammer to crack a nut. The big skill needed is maybe how to decide which approach is needed - simple or full AI, to sort out problems and run complex things. 

Some see AI changing everything, with US Vice president Vance seeming to want to let it rip and roar ahead, without too much regulatory interference. Not everyone will agree with that, but certainly AI is likely to change many aspects of work- and life. Especially if advanced General Intelligence systems can be made to work. However, it may be wise no to get too carried away. What we are talking about here is what is sometimes called ‘machine learning’, with computers being taught to use data to test out /extrapolate patterns. But it is still not really 'thinking'- just advanced pattern recognition. And it’s not the same as human judgement, gut reactions or intuition- or tacit knowledge. That it seems can’t be taught. Or can it? Some say that needs experience, but it might be possible to supply the necessary leaning, or even create it, artificially. All very speculative…

Ultimately though, to round off this bit of speculation, what are we really aiming for? Some of the world’s eco-problems are really complex. Should we leave them up to some new very big AI system to solve, and also to run things optimally?  If so, then we may all be out of work! Maybe not such a bad thing though long term, depending on how it is done, as I explored in a book chapter a while back on Technology and Future of Work. Although would we really be happy handing political and ethical analysis and even maybe decisions, to big AI?  And what about human creativity! Obviously, there are lots of big questions about the future of technology, energy and skills- and big AI.  Maybe you should get an AI chat system to answer them?! Or at least see what it says about it. Over to you! 


Comments

Popular posts from this blog

The IEA set out a way ahead

The International Energy Agency's new Global Energy Roadmap sets a pathway to net zero carbon by 2050, with, by 2040, the global electricity sector reaching net-zero emissions. It wants no investment in new fossil fuel supply projects, and no further final investment decisions for new unabated coal plants. And by 2035, it calls for no sales of new internal combustion engine passenger cars. Instead it looks to ‘the immediate and massive deployment of all available clean and efficient energy technologies, combined with a major global push to accelerate innovation’.  The pathway calls for annual additions of solar PV to reach 630 GW by 2030, and those of wind power to reach 390 GW. All in, this is four times the record level set in 2020. By 2050 it wants about 24,000 GW of wind and solar to be in place. A major push to increase energy efficiency is also seen as essential, with the global rate of energy efficiency improvements averaging 4% a year through 2030, about three times the av...

Nuclear- not good vibrations in France

France is having problems with nuclear power.  It was once the poster child for nuclear energy, which, after a rapid government funded build-up in the1980s based on standard Westinghouse Pressurised-water Reactor (PWR) designs, at one point supplied around 75% of its power, with over 50 reactors running around the country. Mass deployment of similar designs meant that there were economies of scale and given that it was a state-run programme, the government could supply low-cost funding and power could be supplied to consumers relatively cheaply. But the plants are now getting old, and there has been a long running debate over what to do to replace them: it will be expensive given the changed energy market, with cheaper alternatives emerging. At one stage, after the Fukushima disaster in Japan in 2011, it was proposed by the socialist government to limit nuclear to supplying just 50% of French power by 2025, with renewables to be ramped up.  That began to look quite sensible wh...

Nuclear Reliability- an uncertain route

 Nuclear energy provides reliable, baseload, low-carbon electricity that complements the variability of wind and solar’. That, boiled down, is the UK governments view, as relayed in a response by the Department of Energy Security and New Zero to a critique by Prof Steve Thomas and Paul Dorfman. Well, none if it holds up to examination. Low carbon? Not if you include uranium mining, waste handling and plant decommissioning. Baseload? A dodgy idea!  A Department of Energy minister had previously admitted that ‘although some power plants are referred to as baseload generators, there is no formal definition of this term’ and the Department ‘does not place requirements on generation from particular technologies’.  A key point is that nuclear plants are not that reliable- if nothing else, they have to be shut down occasionally for maintenance and refuelling. Add to that unplanned outages, and nuclear plants are not very sensible as backup - especially given their high capital ...