Rolls-Royce intros ‘breakthrough’ framework for ethical AI in Industry 4.0 – and 5.0
Rolls-Royce is to make available new “breakthrough” research on how to construct an ethical framework for the use of artificial intelligence (AI) that people will “genuinely trust”. The company has claimed two templates for ethical industrial AI, which will help AI gain wider trust in society, and accelerate not just the burgeoning Industry 4.0 market, but the future Industry 5.0 movement as well.
The research hinges on two “breakthroughs”, the company said. The first breakthrough is an AI ethics framework, which enterprises will be able to follow to ensure their use of AI in critical and non-critical applications is “ethical”. “This is the first time AI ethics for industrial contexts has moved beyond theory and into practical application,” said Rolls-Royce.
The second breakthrough, within the framework of the first, is a five-step process to ensure AI outcomes (“not the algorithms themselves”) can be trusted. “[It] prevents biases from developing in algorithms undetected; with results being constantly monitored, it ensures they are trustworthy.” The same step-guide applies to every usage of AI – “whether for an ethical process such as sifting CVs or a safety-critical business process”.
The ethics framework and its trust process have been peer reviewed by subject matter experts in several big tech firms, as well as experts in the automotive, pharmaceutical, academic and government sectors. Both will be published in full under Creative Commons licence later this year on the Rolls-Royce.com website.
Warren East, chief executive at Rolls-Royce, said: “The current debate about the use of AI is focused on the consumer and the treatment of consumer data. But we believe that what we have created – by dealing with a challenge rooted squarely in the industrial application of AI – will help not only with the application of AI in other industries but far more widely.”
He added: “We want to move the AI ethics conversation forwards from discussing concepts and guidelines, to accelerating the process of applying it ethically. There is no practical reason why trust in AI cannot be created now. And it’s only with the acceptance and permission of our society – based on that trust – that the full benefits of AI can be realised, and it can take its place as a partner in our lives and work.”
Rolls-Royce monitors up to 8,000 flights every day, it said. It has the capacity, via its analytics-driven engine health monitoring system, to monitor up to 3,000 engines in the sky at one time. It takes about five million measurements per day, from sensors on its engines. AI algorithms scan for system changes and anomalies,, and provide insights to engineers to dispatch maintenance crews on the ground in case anything shows up.
The company’s data innovation team, R2Data Labs, also applies AI to manage risk in the company’s supply chains, to predict demand and improve efficiency. The same is used in its microgrids, created by its Power Systems division, to make its industrial power technology more reliable. The company said it is looking to increase its use of cloud services, and of AI in the design and operation of its products.
The new ethics framework and trust process, developed by R2Data Labs, has come out of a 12 month investigation to demonstrate, specifically, the efficacy of robot inspections of critical components. The peer review process made clear both of these broke new ground, and could be applied to the use of AI across the industrial sector.
In a speech at London Tech Week, last week, Rolls-Royce repeated the mantra that safety comes first. “A safety mindset is fundamental to our products and services, and so it must be fundamental to our data innovation,” said East. It also repeated that AI is critical to its corporate strategy and ambition. “While our aspiration is to become the world’s leading industrial technology company, the impact of our work extends to all uses of AI.”
But the ‘safety’ of AI algorithms, in the broadest sense, is also one of the “biggest public relations hurdles” for Industry 4.0 movement. Rolls-Royce talked about its work in the context of a future Industry 5.0 movement, as well, quoting a forecast from the World Economic Forum, that nearly twice as many jobs will be created by AI automation by 2022 than will be lost because of it.
East said: “From an industrial point of view, we know this is only part of the solution to how to apply AI with trust and within an ethical framework, but it is a big step in moving us and the industrial world towards industry 5.0, where AI is applied more broadly and creates a shift in the contribution that humans make to industrial processes.”
The standard ‘black box’ approach towards AI safety and ethics, to investigate after the event, is wrongheaded, said East. “[It] requires a business to reveal what is often sensitive and proprietary work. It’s also a hard thing to do for a regulator and perhaps… it only provides a snapshot of what an algorithm that is learning, developing, and adapting to its task, is doing at one particular point in time.”
He added: “Our simple, five-step process focuses on validating the outputs of algorithms so there is no need to crack open the black box. It can be applied now, at comparatively little cost, and with great efficacy because the checks can be repeated every few minutes as the algorithms evolves. We are not seeking a commercial return from this work beyond that which its use will achieve for our business and customers.”
East went on: “Certainly, in wider society AI can scare people, and there are scaremongers out there. Like most of the other tools we use, if misused of course there is a dark side. But I believe we can do better than that, dismissing a new technology or tool because of the scary bits won’t help, and isn’t justified. Instead focus on the good bits and it can enable real scale.
“Then we unlock human potential, and accelerate greater industrial, commercial and societal endeavours. Of course we need to reassure about the good bits versus [the] bad bits. Our communities and colleagues must be able to really trust artificial intelligence and believe that it’s being applied for good outcomes; and this is at the heart of the work we have done at Rolls-Royce.”