The data revolution: Transforming products and manufacturing

By: Pawan Bhageria, Senior Vice President - Manufacturing & IT Practice, Tata Technologies

“Our car is fully connected and gets managed remotely”; “Our tractor ‘talks’ to you directly about its maintenance needs”; “We can predict which machine on the shop floor will breakdown and prevent any unplanned downtime”; “Our plant manager takes critical operational actions from her mobile device”; “While on the move, our supply chain expert can keep a tab on each piece of consignment”. These are not lines from a futuristic movie—this is today’s reality. While the data revolution is taking the world by storm, many businesses are quietly harnessing its power to upgrade their offerings and processes.

The result—uniquely differentiated products, never-seen-before customer engagement levels as well as productivity levels, things that were distant dreams until yesterday. You think your business is too mainstream or commoditized to be affected by a revolution? Think again, then. Because chances are that your competitor has already started to act along those lines. Who knows, this might be your chance to be the first mover and reap the benefits of bringing the power of data to your industry.

THE FOUR-LAYERED PROCESS

The implementations of the above kind are usually seen to have a four-layered structure—the first layer consists of various data collection techniques, for example, sensors capturing product performance, sensors within production machinery, measuring real-time status of inventory in supply chain, etc. The second layer is the communication layer between the first and the third. The third layer is analytics, which involves algorithms to process the data that is collected to arrive at conclusions. The location of this processing could be in the product itself or on a cloud computing platform. Finally, we have the fourth layer—the ‘action’ layer that presents the result of the analysis on a dash board to the manager. It also sends an SMS/voice mail to the user and a signal to a machine to perform a certain task.

Let me elaborate with an example of predictive analytics the way it applies to tractors. By embedding sensors in a tractor, one can measure tire pressure—a key parameter that affects fuel consumption of the product. These sensors collect pressure data every two hours and transmit that information through mobile device to a server. The server runs an algorithm to compare standard tire pressure values against the data received and then concludes that the tire pressure is low. Finally, this server sends an SMS to the tractor owner—‘Please check tire pressure on tractor XYZ.’ This simple upgrade increases the lifetime value of the product—the fuel, random pressure checks, and on the downtime that might have resulted if the tire had burst. In other words, the product is differentiated, in terms of convenience. So, in this case the customer would be more than willing to pay a premium for the product.

THE BUSINESS OF ANALYTICS

Many manufacturers have already tasted success in predictive maintenance and are craving more. A multinational chipmaker applied predictive analytics on a single line of core processors process to save $3 mn in manufacturing costs. Flush with success, the company is now looking to expand the scope of predictive analytics to more chip lines to save an additional $30 mn. Sensor-based maintenance management gives manufacturers a rough estimate of how much maintenance service they should plan for and by when in the future.

On the shop floor, it’s important to know which machine is performing to potential and which isn’t and probably requires maintenance. Sensors embedded in an engine can tell us how the moving parts are interacting with each other, help discern patterns in their precision and coordination, and foretell chances of failure. Take vibration analysis. This can uncover common mechanical flaws like component wear and tear, signs of loose bearings, and coupling and shaft misalignment. Such real-time feedback will enable corrective actions on the shop floor. The applications to manufacturing are manifold and only limited by one’s imagination.

Even beyond manufacturing, this technology is making inroads in the after-market services as was elaborated in the tractor example earlier and in supply chain management. Businesses are busy reforming their traditional supply chain by implanting it with sensors and enabling connectivity to support exchange of data with manufacturers across any device or network (public or private company network).

For instance, if a consignment of raw materials moving along a smart supply chain of this kind is likely to reach late at the production facility, the manufacturer has continuous visibility of this. Production heads know where the shipment is located and are in a position to alert customers in time. Besides, they can make necessary adjustments to avoid shutting down the entire production line. Car makers, for instance, are leveraging predictive tools to track how customer orders are moving up the production line, so they can keep customers posted about the delivery date.

Another technology which is redefining business processes and enabling manufacturing managers to monitor and control the operations on move is mobility. This is very powerful and is changing the way supply chain, manufacturing, sales and aftersales operations are

run—making them more agile and efficient. One common concern around this technology is affordability. After all, implementation of this kind of technology requires additional support from trained personnel, expert advice, and infrastructure. One must remember however that not every business needs a full-scale transformation.

Depending on the needs of a particular business, transformations in the range of ten lakh to a couple of million dollars are possible. The returns on this investment can be manifold. The success of these business upgrades depends on a number of factors—the support of the company’s management, expert guidance for implementation, and suitable personnel training to adapt and exploit the technology fully. While this may seem like a daunting

proposition, services are now available to provide start to finish support to ensure success. A third of all data in our world comes out of the manufacturing business and this will only increase in the future. The business is significantly sensor-packed, which makes it easier to

capture many process parameters (eg, temperature of a part being made, stability of machinery, etc). These terabytes of mashed-up data (from internal and external

sources) are an asset waiting to be exploited. At any given time, unlike humans, an algorithm can crack sensor data in thousands. By doing so, analytics gives manufacturers

the deeper and more accurate understanding they need to improve product quality, control operating costs, and make manufacturing more efficient overall.

WHERE DOES DIGITIZATION STAND

In spite of the obvious advantages, only 13% of industry executives polled in a McKinsey survey; released in March this year, rate their organizations’ digital capability as ‘high’. At the moment, digitization is taking baby steps that involve use of digital manufacturing, manufacturing execution systems (MESs), cloud, mobility, social, and IoT. There is more to digitization like simulation and robotics, product styling, product validation, real life simulation in 3D and even 5D! At present digitization thrives in pockets and that too only in some progressive companies. Very few companies have a long-term roadmap for digitization.

Perhaps the biggest hurdle on the road to industry-wide exploitation of these technologies is the lack of awareness and understanding of the technology and, more importantly, usage scenarios. Businesses must start asking questions like—“Digital—what’s there in it for me?” before a competitor brings in a disruptive innovation to their industry on the back of the same technology. For those who are willing to lead the change, companies must start by creating focused teams in manufacturing and supply chain areas, mandated with identifying specific business problems that digitization can solve. Since every business is unique in its own right, this technology can provide maximum returns only when it is strictly tailored to the needs of each business. Companies must therefore work out a digitization roadmap taking along their own unique ecosystem of suppliers, clients, employees and competitor relationships.

Finally let me add this, lest I miss the somewhat obvious: Above all, manufacturing needs to invite and cultivate more people who are passionate about digitization and have a keen understanding of manufacturing. These will be the people who will lead the much-needed change in the industry. And it’s about time.