Advanced operating models
Sep 10, 2018

Autonomous manufacturing: Reinventing factories

Why smart companies are making their shop floors smarter than ever with autonomous manufacturing

Recently I met an old friend at the airport. He's an up-and-coming shoe manufacturer in India. Over the last two decades, I've witnessed – and at times been an integral part of – transformative changes in manufacturing, so naturally we started talking about factories.

I told him about Adidas's SPEEDFACTORY, where customers choose the raw materials, colors, and designs of their sneakers, and have the end product shipped within 24 hours. I also spoke in detail about GE's Factory 3.0 – also known as "The Brilliant Factory." There, the company builds widely diverse products under the same roof with the support of a super flexible supply chain ecosystem. I explained that both factories have undertaken autonomous manufacturing, which is the name for systems that adapt and learn in real time.

His reaction ran between absolute disbelief and nuanced cynicism about a “smart future." Soon, he left for his destination, leaving me bewildered. I wondered why many people like him still regard autonomous manufacturing as a myth. Perhaps my friend didn't know about it because its penetration is still limited. The use of AI in the services industry is more visible – and so more people accept it. By contrast, in manufacturing, many people still regard factories as rigid, human-led operations with fixed processes.

Yet that's not the case in progressive manufacturing enterprises. With the advent of the industrial internet of things (IIoT) and AI-powered neural networks, the 4th Industrial Revolution has the disruptive power to radically transform manufacturing. Simulation, 3D printing, and robotics are some of the innovative technologies driving this digitisation of manufacturing and enabling intelligent manufacturing units.

The encounter with my friend made me question how we can make autonomous manufacturing more ubiquitous across all sectors – large and small.

How does this machine-led ecosystem really work?

Perhaps understanding some of the technology involved in digital transformation in manufacturing is a good place to start. Take AI, for example, which is already taking on various manufacturing roles that previously required human judgment and controls. At each link in the production and supply chains, tools and workstations can communicate via the internet and virtual networks, allowing for real-time adaptations to reflect demand. The replacement of fixed conveyor systems with automated guided vehicles (AGVs) even lets plants reconfigure the flow of products and components seamlessly between different workstations. That means companies can complete manufacturing sequences with entirely different process steps in a fully automated fashion – and they can do so without human intervention. Genpact's team of AI experts has already designed and delivered transformation roadmaps using these technologies for global giants in steel, power, oil and gas, and retail.

Example 1: AI-enabled autonomous manufacturing

The beauty of these innovations is that they aren't limited to production alone. They also transform quality control and maintenance. Historically, factories haven't had a machine health monitoring system or an accurate failure prediction mechanism. That led to frequent unplanned stoppages and inefficient maintenance.

Today, however, sensors and big data analytics are revolutionizing maintenance. That translates to potentially huge productivity gains and better returns for industries such as power or oil and gas, where equipment failure and maintenance often drive up operational costs. AI algorithms can now diagnose causes of failure, decide on the scope of maintenance, and procure the parts and tools needed to complete the maintenance cycle – all without human intervention. That reduces downtime and significantly cuts the cost of maintenance.

Example 2: An autonomous self-learning ai-driven steel mill fulfills custom orders without any human intervention

And that's not all. The sensor revolution, together with big data analytics, can increase the accuracy of drilling decisions to 90%. Prognostic algorithms can recommend the right angle and altitude of drills to maximize the probability of hitting shale deposits.

Example 3: Transforming an integrated global oil and gas company

About the author

Sudhanshu Singh

Sudhanshu Singh

Senior Vice President & Global Practice Leader, Analytics & Research

Follow Sudhanshu Singh on LinkedIn