In industry 4.0, the data generated by connected equipment becomes crucial to optimizing operations.
In industry 4.0, the data generated by connected equipment becomes
crucial to optimizing operations. With sensors that detect efficiency levels of each device, managers can know how well it is functioning, predict when it may need to undergo routine maintenance procedures and, most importantly, remain prepared for downtime.
The benefits are not limited to equipment either. A recent global survey from PricewaterhouseCoopers (PwC) that included over 2000 participants in nine major industrial sectors has shown that about 72 percent of the companies expect data and analytics to significantly improve customer relationships and customer intelligence during the product lifecycle.
“Greater integration of data between manufacturers and customers can open up new collaboration opportunities,” PwC said in a report following the survey. “Clever use of pooled data, for example, can allow manufacturers in B2B markets to help customers in value-chain planning, driving efficiencies within the customer’s operations as well as vice versa. Many companies have such collaborative opportunities in sight.”
Several companies we talked to agreed. Heiner Lang, Senior VP for Business Unit Automation and Electrification Solutions at
Bosch Rexroth, goes to the extent of saying that data is the new raw material for tomorrow’s production.
“In the factory of the future, all machinery will be connected both horizontally and vertically,” Lang said. “Components will be equipped with various sensors which will continuously collect operating data. Turning these data into information and knowledge will enable manufacturers to optimize their processes, allowing for higher productivity and flexibility when it comes to small lot sizes. This development is now past its infancy.”
Where data comes into play
At the basic level, the processes that involve big data remain simple. For instance, devices like synchronous servomotors function as sensors and data sources that collect operating data for real-time processing from the source. An interactive communication platform collects, filters, and processes the data from production lines directly onsite. This is the basis for the continuous optimization of all processes, as well as for an increase in availability.
Going into the details of its benefits, Sunil Mehta, GM of Automotive Business Development, Factory Automation and Industrial Division at
Mitsubishi Electric India, said that in the four-wheelers automotive industries, takt times are less than one minute (for small cars). Hence, big data is generated during the manufacturing and assembly process of each vehicle.
“As the emissions norms become more stringent, it is important that all the engine related data be captured and stored for future needs (traceability function),” Mehta said. “In the future, all the manufacturing processes will result in producing big data and it will depend upon the manufacturer — how to capture and analyze this data. This will also demand secure storage of the data either on the cloud or in the premises. There will also be a need for many data analysts who will work upon this huge amount of data, develop programs to extract necessary data for improvement of quality and productivity.”
But there is a catch
These advantages that Mehta predicts are also turning out to be challenges at some levels. For instance, concerns regarding the secure storage of data that is generated continue to remain unresolved. The PwC report also points out other major challenges such as lack of sufficient talent and old, cumbersome infrastructure. The volume of data generated will increase at a rapid pace as the number of sensors goes up. To store this without taking up massive amounts of physical space, the industry will need smarter compression and storage solutions.
Artificial intelligence and machine learning algorithms will have to play a major role in solving these challenges. Neural networks are not new to the industry. In a recent post, Siemens noted that such technology has been part of steel mills from the 90s. The focus now is deep learning and reinforcement learning. In the coming years, we would definitely be seeing these play a larger role in enabling industry automation.