Capturing industry’s white whale

Data is the backbone of the modern industrial revolution happening around us. Despite its importance, a recent McKinsey survey identified that businesses still approach data with a blinkered view or short-term initiative that fails to make effective use of it. Here, Sean Robinson, service leader at industrial data specialist Novotek UK and Ireland, explains why many strategies miss the mark, and how businesses can capture the elusive white whale that is an effective data strategy.

In Herman Melville’s Moby Dick, we find one of the earliest recorded references of a multitool: a “Sheffield contrivance”, containing numerous tools in the exterior of a pocketknife. In the story, the item serves as a metaphor for the multitalented carpenter responsible for maintaining the ship in its pursuit of the titular white whale.

The metaphor is equally comparable to a good industrial data system, which plays a critical role in the pursuit of an effective data strategy — industry’s own white whale.

In its recent global survey, McKinsey Global survey found more businesses agree that data and analytics are changing their industries in increasingly significant ways. Despite this, McKinsey also noted that many companies are “responding to these shifts with ad hoc initiatives and one-off actions, rather than through long-term strategic adjustments”.

Part of the challenge is that each department of many companies, from procurement and C-suite to field service engineers, still gravitate towards a silo mentality. Yet in order to be effective, the data strategy should run throughout every level of a business, with the capacity for each area to have their own view of the data. If this isn’t considered, the strategy will never truly work.

However, before we determine how to improve a data strategy, we should define what an effective data strategy is.

Defining good data strategy

An effective strategy is one where collection and aggregation of data is supported for every process, practice and piece of machinery that is pertinent to operations. From this data, a company can perform suitable analysis to scale granular information in such a way that it provides macro insights. These can be viewed by, and reported to, any stakeholders with an interest in an area of operation.

As an example, a chief executive for a multinational company might have a top-level view of plant performance, with a simple traffic light system to indicate normal, caution, or warning levels of overall plant health. This is enough for the executive to monitor overall operations and satisfy their interests.

If we assume that one plant in the UK indicates caution, the executive can speak directly with the plant manager. The manager’s view of data will identify specific areas where there is a performance risk such as a specific engine on a particular production line, and have a quick view of relevant KPIs such as availability and uptime as a percentage.

Maintenance staff can view the specific technical issue using the same data at a more granular level, such as engine load information or oil temperature figures, and address it accordingly. If a replacement part or engine is needed, procurement and finance teams can view the make and model and process an order in their system.

This is the white whale, where the flow of data allows different stakeholders and teams to have different views and systems for interacting with one central pool of data. It’s something that has been made theoretically possible by the advent of the Industrial Internet of Things (IIoT), but also hindered by many systems that have been dubbed “IIoT solutions”.

Collecting and scaling

Many IIoT solutions encourage engineers to embrace a simplified approach to data collection, such as only state change data or certain parameter deviations. With this, it is impossible to collect the richness of data required to support several different improvement processes, from fixing the issue to understanding why an engine broke and predicting future breakages.

The best choice is a modern Historian software that can collect various types of data such as state change, alarms, quality system information and process parameters. This allows for very granular insights into specific issues, but it can easily integrate into other software, such as a manufacturing execution system (MES), and run calculations on data sets to provide higher level insight.

It’s here that an effective data strategy makes an immediate difference. One issue that we at Novotek encounter all too often is that businesses employ several systems to collect the same data without realising, and are reluctant to invest in multiple analysis and visualisation tools, such as an MES alongside a Historian software or an IIoT solution.

Fundamentally, businesses are spending money in the wrong places and are subsequently not getting the return on the investment that they want.

The best advice is to think of data like the “Sheffield contrivance” multitool alluded to in Moby Dick, or the modern Swiss army knife. It’s most practical and efficient to use one system with various tools that are applicable to other people, teams and situations. Granular data collection systems can aggregate upwards to add meaning to people at every level, whereas systems that process only aggregated data cannot easily scale down for specific details.

The good news for industrial businesses is that with this in mind, and the white whale in view, it becomes a much less complicated process to capture it. You should set your industrial strategy as one that, for the first steps, endeavours to streamline the number of systems collecting the same data. Instead of a wide array of IIoT solutions, look instead to one store-all system like Historian software.

Then, look at ways to analyse and present the micro-level, granular data at a macro level to serve different business areas. For example, consider the calculations run to present information differently and yield alternate insights. If you are not thinking about who will use the data or have an interest in the data, then your strategy will not be right. Put more than one tool in place if necessary but look at how to share data between systems.

As an industrial data specialist active across the manufacturing, processing, power and utilities sectors, Novotek has extensive experience guiding plant managers and executives alike in the right direction to improve their data strategies. Beyond noting how data can present different insights, the most frequent advice we also give is to not think of system investment as an afterthought of strategy. Incorporating these options into the strategy means you can identify where multiple systems are complementary and beneficial, or where they are a wasted investment.

Data, and the value it possesses, can provide numerous short-term benefits but the biggest advantage comes from a long-term, scalable strategy. With the right approach, underpinned by the right software and systems for collection and analysis, industry can at last catch and conquer the white whale.