The Supply Chain and the Downside of Data

The abundance of raw data-processing power and the advancement of cloud storage have enabled information to be captured as never before. Proponents of Big Data promise unparalleled insight into customer behavior and a seamless flow of information across supply chain networks. But data also has the potential to paralyze businesses processes and encumber decision-making, some experts believe. Companies must learn to manage the volume and complexity of data to truly reap its benefits.

The supply chain is a classic example of an “embarrassment of data riches," according to research from EY. The quantity of data produced by supply chains and their digital ecosystems is not only overwhelming—it has the potential to harm by adding a counterproductive level of complexity that leads to chaos. “Let’s take the example of an EMS model,” said Dave Padmos, EY Global Technology Sector Leader, Advisory Services, “you have all sorts of information pertaining to the EMS that should be communicated in real time. If your customer makes a design change, you need information from purchasing agreements and purchasing contracts for things such as pricing and an approved vendor list; and all of this needs to be automated.” Few companies understand that they have to manage and simplify data before it becomes useful.

The problem is the overwhelming volume of data companies are dealing with, Padmos said. “We are talking zettabytes [trillions of gigabytes],” he said. “Technology has made it easy to capture data for data's sake. The data that is actually useful to a business is obfuscated by the amount of data it has accrued, or that it is not even aware of.”

Align data with strategy

The first step toward avoiding chaos is pretty simple – “Stop and ask yourself what you are trying to do with your data?” said Padmos. “What is your data strategy and how does it support your business goals?” For some companies, he said, data enhances the business process. For others, the data itself is the goal. “If you look at sharing-economy businesses,” he said, “all they have is data.” In other cases, data analysis improves productivity – such as an EMS receiving component and pricing information along with an ECO.

The rest of the transformation is not so easy. EY recommends companies establish an enterprise-wide data management strategy before they attempt a digital transformation. “Instead,” said Padmos, “companies are capturing all the data they can.” Business can start with managing data within the four walls of the enterprise. If a company is currently using multiple incompatible systems, a single data analytics platform can help aggregate data. “Getting your own house in order is never a bad answer,” said Padmos.

Most critical for the supply chain is a strategy that encompasses an integrated view of how data flows through the network and an understanding of how data drives productivity. In an ecosystem such as a manufacturing, supply chain information has to be standardized and automated. “How can you do an analysis on an ECO if information comes in as an attachment to an e-mail?” Padmos points out. Systems also have to integrated, he added.

Machine learning, et al

The supply chain can benefit from advanced technologies such as machine learning, according to EY, although it is not a push-button solution. To be most effective, data analytics and machine learning require companies gather enterprise data, establish an enterprise data management strategy, and direct systems based on driving business goals. When properly directed, machine learning can excel at classification of unstructured data and matching similar data from disparate environments. This is the kind of end-to-end visibility envisioned by the supply chain. However, EY estimates that as much as 80 recent of a large enterprises supply chain data is likely in the hands of other companies in its external ecosystem of partners.

One company has applied machine learning to the freight industry. ClearMetal’s Predictive Intelligence Platform leverages the cloud, artificial intelligence, and a proprietary network simulation engine to predict container, vessel, and shipper needs with a high level of accuracy. The platform collects granular data across a supply chain network and runs “what if” scenarios against that data. The results can help freight companies make informed decisions about their physical assets. Adam Compain, CEO of predictive analytics company ClearMetal admits that artificial intelligence and machine learning are uncommon terms in the supply chain, but “we’ve been able to see the impact machine learning could have." Freight companies currently rely on past customer behavior to determine how and when to ship whereas AI can help anticipate customer needs. “Carriers and shippers are excited to use the next wave of technology,” Compain said. “The data is already there. They can see how artificial intelligence can be applied more proactively to move from visibility to prediction.”

Padmos and others reiterate the necessity of an enterprise having a data-management and simplification strategy at the beginning of a digital transformation for another reason: the data deluge is only going to get worse. The first thing the IoT will do is add to the barrage of data.

Blockchain technologies promise benefits potentially greater than the cloud-mobile-social-big data technologies that supply chains are grappling with today. Blockchains' ability to automate trust through a distributed digital ledger database and automate transactions when pre-set conditions are expected to significantly increase supply chain efficiency. Blockchain could become a single point of truth—with a since data format—for all partners in an ecosystem, according to EY.

Unmanaged, EY research concludes, data complexity becomes a barrier to innovation. To seize the full potential of the digital transformation, companies must develop data strategies, and better information and data management discipline, and start asking better questions.