“Executive interest in performance metrics is high, yet companies often don’t achieve the results they want.” This is the view of the Manufacturing Enterprise Solutions Association (MESA), and it’s one that we regularly observe in European businesses. Here, Sean Robinson, service leader at Novotek UK & Ireland, explains what manufacturers must do to get the most out of industrial data.
Product and process innovation matters more today than ever before. To set themselves apart, manufacturers must always be working one step ahead of their competitors. For example, due to the large amount of different brands all competing for space on the supermarket shelves in the food industry, manufacturers must constantly innovate.
Own-brand products are becoming increasingly popular across Europe, with well-known brands carrying out large-scale promotions in 2017 to regain favour with consumers. However, own brands are fighting back by investing in innovation. Gerald Lindinger-Pesendorf of German consultants Dr Wieselhuber & Partner said, “up until recently, private labels mainly copied big brands, now the best ones are innovating”.
While important for the business, innovation brings challenges for plant managers, who must change or add processes in the production line. Plant managers must also meet increasingly strict supply schedules to get the products to the shelf on time. They must also balance this with reducing costs, as the own-brand products must still be cheaper than the branded goods.
To be constantly adaptable and cost-efficient, manufacturers in all industries, not just the food sector, need to have an accurate representation of what is going on at every level of their plant. This will allow them to see where improvements can be made.
As the MESA International’s Metrics that Matter report says, “metrics matter in business performance. The better the company’s system for metrics, the more their operations performance improves, and the more their business and financial performance improves.”
There is a multitude of data for plant managers to measure, especially as connected devices are becoming increasingly common in most plants. Sensors are either inbuilt or retrofitted to most machines, collecting granular data such as output from each machine or motor performance.
This granular data should be fed into wider metrics such as Overall Equipment Effectiveness (OEE) or Right-First-Time quality, or submitted into functional systems that determine when inventory should be dispatched. However, this is where 95% of manufacturers fail.
What mistakes are made?
The biggest problem that we encounter when approaching customers from all manufacturing sectors is that they are unable to see the problems in their current data collection systems.
Board level staff often believe that their existing ERP systems are sufficient to monitor their plant. They may monitor production as an input and output, potentially even dividing the process into a couple of cost centres. However, as plant managers will know, the production process can have numerous stages, all of which have their own variables, so a couple of cost centres are not sufficient for accurate reporting.
For example, in a plant producing chips, the potatoes are sorted, washed, cut, conveyed and then packaged. At each stage in the process, granular data needs to be collected to form an overall image of the plant, but an ERP system is incapable of collating such a detailed amount of data.
To monitor a plant correctly, the board will need a control system that can collect the most granular of data. GE’s production management systems collect incredibly detailed data, which can then be scaled up to the level required.
For example, the shift manager may need to understand that one worker was not working at a quick enough rate, or that there was a two-minute stoppage in production due to a technical issue. However, for the plant manager, they may need to see production output over a whole day, with only more significant maintenance issues being flagged up.
The chief operating officer may only want to see production over the month, but each person must have enough relevant data accessible to them to help them be accountable for the output of the plant.
This is another problem that we commonly encounter in manufacturing plants. CEOs and board level employees often use metrics such as OEE, or look at adherences to schedules, but quite often, this is not actually based on sufficient granular data.
We’ve encountered a number of plants where when we question where those figures are originating from, the staff cannot be sure, as the production management system is not collecting all of the granular data from the factory floor. This means that board level executives are making decisions on figures that do not accurately represent what is going on in the factory.
What data should be used?
Once all the granular data is available, it can seem overwhelming to any member of staff looking to base decisions on this data. Therefore, to use the data available in a successful way, the operations team must determine what metrics are truly valuable to the smooth running of their plant.
According to the MESA Metrics that Matter report, “what metrics actually matter to an operation depends on strategy, industry segment, process type and production and market conditions.” Identifying the correct metric could lead to a competitive advantage for a business.
For example, in the food industry, where there are strict regulations and consumer image is important, emphasising the traceability of the supply chain of ingredients could be beneficial. Alternatively, if product or process innovation is important, then if the plant can demonstrate issues in the production process before new products are produced at full volume, this can minimise any wasted product or undelivered orders due to technical difficulties.
Once senior staff have determined the appropriate metrics for them, the business must ensure that they are collecting the right data to assess these metrics. To do this, they must use a data platform that provides a secure path for machine generated and human generated data to land.
The right data platform
The data platform should then be able to share the data across a series of functional systems. In the past, data was siloed into different functional systems, meaning that staff had to work with different systems to assess each metric, rather than being able to see the whole collection of data across the board.
GE’s suite goes a step beyond. Not only is the data shared across the core functional systems such as reliability analysis or material tracking, the suite also offers machine learning tools that can help to determine a solution for difficult consistency, quality or reliability problems.
These advanced analysis tools can discover hidden relationships in manufacturing data, allowing plant managers to learn more about the capabilities of their plants and how they can use the existing structure of the plant to solve the problems that they are facing.
The performance management suite is also open to integration with other systems, for plant managers who need detailed investigation or analysis tools, or work process management functions to improve consistency in the factory.
From the most basic level to this advanced level of data analysis, no matter what the metric is that is being measured, plant managers must be confident that they are collecting the right data at a granular level. This means that all machines and processes should be monitored closely, from motor efficiency to production outputs.
Senior staff must ensure that this data does not go to waste by investing in a data platform that spans HMI/SCADA, manufacturing execution systems (MES) and analytics. By doing this, they will be able to turn the data into actionable information.
Many companies must present executive level staff with metrics to enable them to make decisions about business strategy. However, these metrics must be based on accurate information, generated from a granular level and the information must serve the needs of the person receiving it and the needs of the business. Otherwise, no matter how much interest there is in metrics, companies will never be able to improve their processes and achieve the results they want.
