Technology & Innovation

Automation and the data-driven future of manufacturing

May 05, 2016

North America

May 05, 2016

North America
Our Editors

The Economist Intelligence Unit

_____________________

Why the growing sophistication of automated manufacturing accentuates the strategic value of data

Sponsored By
 

Ever since Henry Ford designed and implemented the first moving assembly line in 1913, automation has been the driving force in the evolution of manufacturing. As this automation has become increasingly computerised, managing data has emerged as an essential capability for manufacturers.

Today, the sophistication of manufacturing automation is advancing rapidly. As a result, the effective use of data is not just improving efficiency but is also driving greater agility and deeper integration with supply-chain partners—and is even giving rise to entirely new business models. This will only increase the requirement for information-management capabilities for manufacturing firms.

One area where manufacturers are applying increasingly sophisticated automation is in boosting quality. It is estimated that as much as 30% of a manufacturer’s annual revenue can be lost as a result of quality defects in production, so anything that can substantially reduce the number of defects will obviously have a noticeable impact on the bottom line. Manufacturers are therefore collecting more measurements from a larger number of instruments, and they are doing so more frequently in order to spot early indicators of quality issues as soon as they arise.

In 2012, for example, chip maker Intel revealed that it had saved US$3m by analysing the data produced by its manufacturing equipment for one line of chips to predict quality issues. This also allowed the company to greatly reduce the number of quality tests it puts its products through, thereby accelerating production. In order to achieve this, however, Intel had to process 5 terabytes of machine data per hour.

The next step—beyond simply predicting quality issues—is feeding those predictions back into automated production systems so that they can be resolved without human intervention. This means that production processes can be constantly tweaked, creating a process of iterative improvement that benefits both quality and yield.

The opportunities for intelligently integrating automated systems to improve manufacturing stretch beyond the walls of the factory into the supply chain. The result is often described as Industry 4.0—an ecosystem of Internet-connected “smart” factories which communicate with each other and with logistics providers. This integration can even stretch as far as the customer, either directly, via the Internet, or indirectly, via data collected from sensors embedded in products.

One business at the forefront of this networked model manufacturing is US company Local Motors, which is preparing to launch the world’s first commercially available 3D-printed car, produced in 18 “micro-factories” situated near cities. By using 3D printers located near its target markets, Local Motors not only hopes to dramatically reduce distribution costs, but also to allow customers to personalise their cars automatically through a simple online interface.

As these examples show, data lie at the heart of the evolution of manufacturing automation. But the growth of data also presents a number of challenges. Indeed, the challenges that manufacturers face in digitising their operations are arguably greater than in many other industries.

One challenge is the sheer volume of data. A single machine’s performance logs may dispense around 5 gigabytes of data per week. But a greater challenge is the variety—manufacturers typically use a wide range of systems, from an array of suppliers, which produce both structured and unstructured data. This can include anything from streaming data from manufacturing systems to video feeds from CCTV cameras to shift reports from human operators. Much of this is semi-structured: not easily tabulated and conforming to formal database structures but tagged or marked up to allow it to be managed.

Automation offers manufacturers considerable benefits, but it also presents companies with an information-management challenge that they must overcome before those benefits are realised. Manufacturing firms’ future success will be determined as much by their ability to wrangle data as by the quality of their products.

Enjoy in-depth insights and expert analysis - subscribe to our Perspectives newsletter, delivered every week