READ FULL ARTICLE HERE: "Promise and peril as engineering firms roll out AI"
Asked as part of the recent Professional Engineering and IMechE survey on the introduction of AI in engineering, the question’s answers revealed some major concerns – and some opportunities for innovation. Here are some of the most insightful replies, organised by respondents’ sector of work.
Environment and sustainability
“AI lacks ethics and judgement in its application, and thus has no real value. It may be good at finding needles in statistical haystacks, but not much more. It is not a panacea for the salary-slave overworked society.”
“There is always a risk in automating that people will lose basic skills and understanding of why designs are done in a certain way. Errors are also likely to increase, as AI has [been] shown to create ‘averaged’ results and not necessarily bespoke to the specific needs, which may be out of range.”
“The main risk is over-reliance on systems that are seemingly ‘black-boxes’. Engineers need to be able to understand how the systems work and check that the outputs are reasonable.”
Power and energy
“There is a real benefit to automating mundane tasks. The main risk is the ‘black box effect’: AI gives an answer, but the reasoning may not be transparent.”
“AI is only as good as the program designers. The risk is that AI will be introduced without any human interaction from engineers experienced in engineering techniques. Also, younger engineers will believe what the model says, forgetting to question and challenge. I despair for [the] future [of] engineering.”
“The biggest risk is trying to introduce AI in areas of engineering where it should not be utilised. The biggest opportunity is where AI can contribute to making better engineering decisions, increase productivity, and enabling engineering to better the lives of society.”
“AI, like many new tools/approaches, will take time to mature. We’ve seen this with AI-generated art, which has rapidly improved from crude, inaccurate pieces to images which are challenging to discern from reality. There is a risk that AI dominates and that the race for efficiency displaces originality, along with the skills that it displaces. We’ve already seen an element of skill loss through ‘traditional’ automation, but how many people want to be tinkering with carburettors any more?”
“The main risk is people spending time and money on researching AI usage that never amounts to anything and cannot [add] value. This will happen if people do not understand the technology. On the plus side, it might help us identify redundant tasks, since if a task can be [completed] fine by an AI, it could either have been eliminated by better process/work design or never added value in the first place.”
Aerospace
“Using AI for generative design or other decision-making tasks feels incredibly dangerous due to [the] lack of traceability and accountability.”
“I think AI may be a powerful tool, but we need to consider carefully how best and where best to apply it. There is no substitute for human creativity or ingenuity.”
“There are both risks and opportunities. The pace of the technology itself is a big challenge. Universities need to react quickly to this but I worry they will struggle to adapt fast enough. AI will have an impact in particular on lower-level repetitive tasks, leading to less manpower being required. AI will also allow more and deeper analyses to be carried out, which will lead to new products and services, and possibly new jobs.”
“[I am] nervous about what the future holds. Potential for disruptive technology? Will we rely on AI and lose the fundamental basics of engineering that are required to verify and validate designs?”
Automotive
“I think there are opportunities, particularly machine learning and generative AI, which will help. A computer should be able to more easily and more quickly identify patterns and check against known problems. On the other hand, human nature will encourage people to blindly believe the results of any AI task, which could be a problem.”
“There will be a temptation to trust the results/outputs, as with other lower-level tools we already have. Human calculators that previous generations had have completely been replaced by computer software tools, but the users still need to be experienced and disciplined to reality check the results/outputs. The number of people employed to do the lesser-skilled work has decreased, I am sure, but the engineers now have to understand the nature of the tools used and be the masters of them, not having to rely on their results/outputs blindly.”
“AI gives a totally authoritative answer, however this can be completely wrong! It’s a fast search tool and gives the odd hint, but I hope engineers continue to approach its answers with some critical thinking.”
“[The] main risk is overdependence and associated loss of knowledge, and understanding of core principles.”
Mechatronics, informatics and control
“As with all computer modelling, rubbish in equals rubbish out. It does take a lot of experience to spot invalid assumptions et cetera, and I wonder if there is enough on the internet to train AI how to do this? Reports often only tell you the success stories, they miss out all the learning on intermediate prototypes along the way. Automotives have been made reliable because of 100 years of use in all possible ways and environments, and that experience fed back to the next generation. Without this real-world input, how will AI produce better results?”
Process
“Less oversight/human input to the design process could mean less innovation, but existing design solutions should be better optimised.”
Rail
“Individuals and companies believe everything that the AI is telling them, without undertaking a thorough check of the input and outputs.”
Biomedical
“Because generative AI relies on a large database, it is an expert at taking inspiration from existing designs – too much reliance on AI may stifle innovation.”
Want the best engineering stories delivered straight to your inbox? The Professional Engineering newsletter gives you vital updates on the most cutting-edge engineering and exciting new job opportunities. To sign up, click here.
Content published by Professional Engineering does not necessarily represent the views of the Institution of Mechanical Engineers.