ABB CTO Bazmi Husain has a digital strategy designed to maintain the Swiss engineering giant a leader in industrial automation.
His strategy is focused on two rapidly-changing areas of the business: energy and industry.
"On the energy side, the energy revolution is changing how electricity is generated, how it is transported and how it is going to be consumed. On the industry side, we are going through the fourth industrial revolution," Husain explained at Huawei Connect in Shanghai, China.
"One thing that is clear to us is that for these revolutions to succeed and deliver the value that is expected of them, we have to go beyond automation. Why do we need to do that? Because automation fundamentally has to be told how to react to every situation that it encounters.
"Given the growing complexity that is developing in industry, we need systems that not only react as they are programmed but are able to handle situations that are not planned ... These systems are what we call autonomous systems."
ABB traces its routes to 1883, when Swedish industrialist Ludvig Fredholm founded electrical lighting generator manufacturer Elektriska Aktiebolaget in Stockholm. In 1988 it merged with the Swiss company Brown, Boveri & Cie (BBC) to form the ABB Group.
In its 136-year history, ABB has been responsible for a long list of industrial innovations, including the first steam turbine in Europe, the first combustion gas turbine for generating electricity, the first company in the world to manufacture synthetic diamonds and the first nuclear power plant in Sweden.
ABB has also been a pioneer in for over 40 years. In 1978, three years before Husain joined the company, it launched one of the world's first industrial robots. In 2014, the firm unveiled YuMi, a human-friendly dual-arm robot that ABB calls "the world's first truly collaborative robot".
Husain expects AI to play a central role in ABB's next breakthroughs.
"We do believe that the cornerstone technology for moving towards autonomous systems is going to be artificial intelligence," he said.
Read next: How UK CIOs are using automation
AI has already proven its predictive value in industrial applications such as condition monitoring and anticipating failures, but it still struggles when applied to unplanned situations.
Difficulties with data are one of the biggest challenges. In industry, modern process plants have distributed control systems that process vast volumes of data from thousands of different types of tools that have little in common, such as transformers or robots. This makes it difficult for analysts to find integrated insights.
They may also have years of data on a device that has never before failed. Fully functioning equipment is not a differentiator in industry – it's a minimum expectation. But if there is no data related to failure, it's difficult to train an AI on how to recognise the risks or find an outlier indicating exceptional performance that can be made standard.
The AI then needs to be deployed on devices, the edge and the cloud, in real-time applications and on constrained equipment. Most chips today don't have the inference times to fulfill this requirement.
Industrial AI also requires algorithms that are explainable and highly reliable.
"Even more so when the consequence of a wrong action can be either very dangerous or expensive – or both," said Husain.
Husain shared his vision for industrial automation through the example of the mining industry, where many processes are not as safe, productive and efficient as they could be.
In the future, he expects a confluence of technologies to take humans out of the dangerous work of mining and into safe control rooms above ground, where they can use digital twins and virtual/augmented reality to see and control the processes.
"It is all integrated so if there is a problem in one segment of the plant they now how to deal with the following segment," said Husain. "In the operations part, humans are replaced by autonomous equipment or moving the ore, for charging the mines, for new blasting areas.
"This is all done autonomously. And then if you have a problem, utilising digital twins and diagnostic capabilities you are able to actually identify and fix the fault much quicker and much better."
Numerous technologies from different companies will have to be brought together to achieve create this.
Husain points to a recent collaboration between ABB and Huawei to explain how this can be work. The companies worked together to create an industrial AI-powered waste separation unit that automates the recycling process.
It comprises a conveyor belt that brings the garbage to the unit, a vision system that detects the presence of the garbage and sends the data to an intelligence unit that classifies the material of the waste, and a robot system that picks up the item and puts it in the right bin.
It took just one month for ABB and Huawei to design and produce this system.
"For this kind of vision to happen, it takes a lot of technologies that have to be combined together," said Husain. "ABB believes that there is no one company that can do it. This has to be done in partnership. And the message that I want to leave you with is that it is partnerships like Huawei and ABB that will make a bridge to this future of industrial autonomy."