Industry 4.0 is a broad church. The term denotes a wide variety of technologies and concepts, from connected factories, to cyber-physical systems to advanced cloud computing. Lending it a precise definition is complicated enough without having to take regional differences into account, such as the varying levels of access to technology and resources, or the unique conditions in individual markets. But if emerging economies are to take advantage of Industry 4.0, these are all essential considerations.
So says Guatam Dutta, Senior Director, Marketing at Siemens Industry Software Solutions. Dutta has responsibility for the Indian region, where automotive growth continues at an impressive pace. OEMs such as Maruti Suzuki, the nation’s largest, are expected to set sales records for FY2018. But whilst Dutta doesn’t believe that an Indian vision for Industry 4.0 differs greatly from its German counterpart, its framework does need to be squared with what he calls an ‘Indian reality’.
In India, he explains, there is an extremely large and mature base of Tier 1 and Tier 2 suppliers, with examples along the entire length of the value chain, from steering and safety supplier Rane Group, to rear-view mirror manufacturer SMR. These suppliers, he says, serve OEMs all over the world, and their manufacturing operations reflect this. Beneath them, however, is where the real difference lies.
Today’s suppliers are being hit by huge challenges on the development front, as they are no longer making single-science products, but multidisciplinary ones which combine electronics, hydraulics, mechanics and software
“There is a huge tail of Tier Three, Four and Five suppliers which need to transform the way they do business, and adopt digital technologies if the Indian automotive environment is ever going to undergo a real digital transformation,” says Dutta. “Compared with markets in the developed world, there are relatively few OEMs building in India. And so mass transformation will not happen if only they and a few hundred suppliers adopt the digital technologies. It needs to be the 10,000 suppliers enabling it all, making anything and everything from sheet metal parts, to castings, to forgings, to small electronics and assemblies – these all need to come together in the automotive value chain.”
In short, the Indian automotive industry bottom heavy, a wide-based pyramid filled with important suppliers who, nevertheless, work with limited resources and under huge price pressure. Simply suggesting an operation ditch its existing set-up and purchase a set of new machines, says Dutta, is not a viable option.
As such, Siemens’ job is to find a way of enabling inputs on dated machinery. “For example,” he says, “there are sensor technologies that can be retrofitted today which will allow current analogue machines to collect digital data, helping owners to make quicker decisions.” But exactly what systems could this data be fed into which, in turn, could revolutionise the country’s automotive sector?
One thing that Dutta is keen to stress is that technology itself is not the most important consideration in Industry 4.0, because technology is always changing. What is more important, he says, is framework, which in Siemens vision is composed of three major components. Each corresponds with the three major parts of a vehicle’s life, namely design, production and use. Digitalisation creates for each of these a digital twin – a virtual copy of a real-world machine or system. Changes in the latter can be immediately mirrored in the former, and simulations can be run within a digital twin to improve efficiency and solve problems.
Essentially, the machine is learning decisions made by human beings, and in turn, it will create and find solutions to its own scenarios using logic from past scenarios
A digital product twin, for example, assists designers by helping them quickly come to terms with customer requirements and bring products to market quickly. This could also include digital validation tools, allowing designers to validate something before its even manufactured.
Of the three twins, digital product twins are the most mature, says Dutta, with the concept having existed for over ten years. However, systems integration, he says, means that digital product twins are more important than ever for suppliers, particularly as time-to-market requirements continue to tighten. “Today’s suppliers are being hit by huge challenges on the development front,” he says, “as they are no longer making single-science products, but multidisciplinary ones which combine electronics, hydraulics, mechanics and software. Today, OEMs want subsystems off their suppliers.”
Suppliers will also benefit from increased use of digital production twins. Even a simple system like a windshield wiper, says Dutta, is suddenly made far more complicated by the inclusion of a feature such as automatic wiping when it starts to rain. The system then requires sensors to detect changes in weather, and software to instruct the wipers how to run. The supplier is then working with perhaps three different teams, including suppliers from the lower tiers, all with different delivery timelines. Digitalisation could synchronise these timelines.
Digital production twins also allow for simulation of a factory line prior to making substantial investments in a system. This assists in detecting faults, and in the future, sensor data on the line can suggest actions to improve efficiency and output. This will be a priority for OEMs, says
For OEMs, mass customisation is one of the big drivers behind digitalisation. One line for one platform is no longer good enough
Dutta, as they are the ones who have the most manufacturing issues, particularly at a time when customers are asking for customisation at mass-manufacturing prices.
“For OEMs, mass customisation is one of the big drivers behind digitalisation,” says Dutta. “One line for one platform is no longer good enough. They need to monetise their capital investment by ensuring infrastructure can manufacture multiple models and variants.” Digitalisation makes this possible by enabling OEMs to explore complex manufacturing layout options.
The final twin is the digital performance twin, which Dutta says will underpin some of the emerging megatrends seen in the automotive industry, which in turn will change the way other real-world businesses operate. “The world is now looking at extreme electrification of vehicles, as well as greater autonomy, and a shift towards shared mobility,” he says, “and so whilst digital product and digital production twins are important, they aren’t enough.
These new technologies on the road need to be supported throughout their lifecycle. This is what the performance twin enables.”
The performance twin could enable an age of what Dutta calls ‘prescriptive maintenance’, as opposed to preventative maintenance. By combining data on the part with data on the driver’s style and usage on the road, a performance twin can schedule maintenance months in advance.
The world is now looking at extreme electrification of vehicles, as well as greater autonomy, and a shift towards shared mobility. These new technologies on the road need to be supported throughout their lifecycle
In turn, the performance twin can signal issues with the production twin. Data gathered on the road can be sent into the cloud, and used by the production twin to improve on existing designs and manufacturing methods.
Leaving the decision making to the machines?
But as Dutta correctly observes, data does not equal knowledge. Data must first become information, and information must be interpreted before it becomes knowledge. To date, this has largely been the job of humans, but artificial intelligence (AI) and machine learning could take on fundamental design and manufacturing decision-making tasks.
“Imagine how many decisions are being taken every day to meet certain requirements in the automotive industry,” says Dutta. “There must be simply millions of combinations, realities and scenarios, and if they were all available digitally, they might require very little redefining in future. A machine could look at a problem and, delving into a digital library, see how it was solved in the past.”
This is the role of machine learning in Industry 4.0, says Dutta. “Essentially, the machine is learning decisions made by human beings,” he explains, “and in turn, it will create and find solutions to its own scenarios using logic from past scenarios.” Improvements in capabilities could prove exponential, and as Dutta concludes, this demonstrates the intrinsic relationship between machine learning and Industry 4.0 – without digitalisation, machines would have neither the data nor the means to receive it to strengthen their performance.
This article appeared in the Q1 2018 issue of Automotive Megatrends Magazine. Follow this link to download the full issue