When a multinational oil company needed to predict equipment failure in its wells, it applied machine learning from Microsoft to existing data, and within four hours was able to identify six wells in need of attention. The company called the sites—and learned that the predictions were so accurate that four of the six wells were already in system failure.
When Swedish packaging and processing company Tetra-Pak wanted to reduce downtime, improve performance and enhance customer service at 100,000 sites in 140 countries around the world, it connected its systems to the cloud. After a six-month trial supporting 11 customer lines, downtime was eliminated by up to 48 hours per line, at a savings of 30,000 Euros per customer.
These are real-world examples of digital transformation in manufacturing. Some businesses are farther along this process than others, but all businesses must make this change in order to flourish—perhaps even to survive—in the current environment. The opportunity is larger than ever. Manufacturers who take advantage of their data have the potential to generate nearly $400 billion in revenue in comparison to their peers who don’t, according to one estimate. Microsoft’s recommended approach: Think big, start small, go fast.
Today, many companies could be considered data rich and information poor; they collect or generate a lot of data but don’t bring it all together to inform their decisions.
For manufacturers today, one key potential of data is to predict where breakdowns in a factory may occur. For manufacturers tomorrow, the power of data is being able to connect the enterprise to the supply chain; predicting breakdowns that could stop the assembly line; servicing disruptions that inconvenience customers; or figuring out the high maintenance costs due to repeat visits. Data will enable the system itself to take action by ordering a new batch, alerting a supplier, stopping a line or performing any number of other tasks. This is the ultimate level of automation, and allows company executives to operate at a strategic altitude within their organization.
"Microsoft’s recommended approach: Think big, start small, go fast"
Within Microsoft’s own Manufacturing and Supply Chain, we have already put the technology to achieve this automation into play. In the first phase of an automated solution that uses Microsoft Power BI, PowerBI Mobile and data analytics to help Microsoft and factory managers streamline operations. As a result, we’ve seen a time savings of 15 to 20 hours per week, time that was previously spent manually building out extensive reports. And more companies join this process every day. By 2022, 40 percent of operational practices will be self-healing and self-learning, according to one prediction.
The data that generates these outcome-focused insights will come from three areas in the manufacturing enterprise. The first is the information from what Microsoft calls the intelligent edge. This might include sensors in a plant, historical data, production line data or employee inputs. There is so much data at the edge or the plant floor, and according to some research, only one percent of it is being used for intelligence. Why not leverage this information?
Second is the information that comes from all the different edge locations, across time zones, plants and geographies. Those insights can be brought together and combined with other data sources to make any enterprise more intelligent as a global, connected entity—forming connections not only in the data itself but across factories and suppliers as well. In doing so, the manufacturing organizations can give team members one holistic and consistent report that they can access from anywhere, saving time on revisions and updates. Finally, when the connection is made back to customers, front-line employees and suppliers, the magic will start.
Large manufacturers are now truly able to optimize their supply chains, operations and asset performance at a global level. Tools that a cloud like Microsoft Azure provides such as virtually unlimited data storage and computing power, the internet of things, advanced analytics, artificial intelligence and advanced digital-twin technologies, can accelerate this transformation. Working together, these tools will connect factories and suppliers from around the globe to further streamline processes and decrease downtime across the board. To get started in this process, a company must identify its business goals, connect the data that can serve those goals, and interpret that data to understand what actions to pursue.