Today, businesses can integrate the upstream and downstream value chain cost-effectively − using digital solutions that enable end-to-end visibility and traceability.
In part 1, these topics were discussed as well as how forward-looking food and beverage manufacturing operations − through horizontal organisational alignment and better data capture and integration − have driven significant business performance.
Now I share my insight on the ways data ownership, management, and sharing drive that end-to-end visibility that manufacturers are looking for.
Also read part 1
Data ownership emerges as a key business driver
To achieve such levels of visibility, some fundamental challenges must be addressed. First, machines must be connected, and the data extracted from those machines. This is easy to say but may be difficult to implement. Why? Most pieces of plant equipment come with proprietary software. Basic operating data can be accessed, but attaining additional data usually requires the equipment supplier’s permission or may even require negotiating a costly services contract.
In today’s digital business environment, success hinges on an organisation’s ability to access and control its own plant operational data at all levels. There should be no restrictions on using asset data to improve business performance. Paying a third party for the right to access plant data no longer makes any sense as a viable business approach.
Food and beverage manufacturing best data practices
Any data model representing a comprehensive view of the food and beverage manufacturing plant’s operational lifecycle should achieve the following fundamental best practice elements.
The plant data should be:
- Reliable and repeatable – What if sensor data is coming to operators and analysts in an uncalibrated manner? This could cause inconsistencies that lead to the wrong operational decision. This is why the reliability of data represents a key building block.
- Comparable – Plant management should be able to make apples-to-apples comparisons across plants. Comparable data can help to determine important differences such as overall equipment effectiveness (OEE) and lay the groundwork for continuous improvement initiatives.
- Trustworthy – In cases where data has been manually entered, the risk of error increases significantly. Accounting for data accuracy is critical in establishing data trust.
- Transparent – Key performance indicators (KPIs) should be analysed as pyramids, with the KPI at the top of the pyramid linked to all cascading KPIs below it. In this way, it becomes clear to plant operators and management how the various processes in the plant support each other.
As food and beverage operations evolve to more digital operations the data ownership issue has become much more significant. In some cases, external organisations such as OEMs claim ownership, while end users identify themselves as the owners in other cases.
These issues are important to sort out because, if left vague and open to question, multiple versions of the truth could emerge when operators are making decisions that directly impact plant profitability.
For this reason, the topic of master data – which dictates who owns which data and where that data resides — should be taken very seriously.
Today, running a plant can be compared to operating a Formula 1 racing car. In designing a reliable car, it becomes important to model and simulate its performance before the car is even built so that performance can be optimised. And then, once built, it becomes critical for the racing team to own the data coming from the car as it operates in a competitive environment.
Robust data management aids the migration of standalone machines to integrated operations
When multiple OEMs are engaged in the operation of a food and beverage manufacturing plant, end users must manage separate service contracts to work with each of the OEMs. Although this can be effective for operating an individual machine, such an approach does not guarantee an integrated and efficient line or plant. When upstream and downstream machines are not synchronised and issues arise, it is not uncommon for finger pointing to occur.
For example, if a line running three SKUs is operating at 85% OEE, who is responsible when, six months later, an additional five SKUs are added and OEE plummets to 40%? Sorting through such issues is both frustrating and time-consuming. That’s why robust data management is an important critical success factor.
Data sharing is now fundamental to driving profitable operations
To remain competitive, the nature of how food and beverage operations manage process data has to change. In the past raw data was reformatted into readable information, which operators often failed to utilise because they trusted their tribal knowledge instead. Later, that data was used to drive more informed decisions, but this intelligence and wisdom was shared only after an avoidable negative situation had occurred.
Today, data is much more critical because it has to be used to drive self-learning (machine learning in many cases), situational awareness, and self-healing through predictive maintenance practices. The nature of the human-to-machine interface has been permanently altered thanks to new ways to rapidly gather and analyse data.
Open platforms that enable the easy sharing and analysis of data represent the future of food and beverage manufacturing. So do object-oriented programming approaches that allow technologists to build products and processes more rapidly and with more accuracy. This high-speed product life cycle data sharing ensures that nothing is lost between developing a product concept and delivering that product and all its associated customer benefits to the marketplace.
Food and beverage organisations that are expediting their migration to truly integrated, data-driven operations are now taking some important fundamental steps. First, they are defining their value chain data model.
Next, they drive the governance needed to ensure plug-and-play compatibility—either today or in the future. Finally, they are assuring that the data they gather and access is rich and contextualised so that an Industry 4.0 level of business intelligence and autonomy can be established and safeguarded.
Article by Michael Jamieson – segment president for the consumer packaged goods industry for Schneider Electric.