Maximising equipment effectiveness through production innovation

• For industrial companies, the gap between intended and actual outcomes of industrial digitalisation innovation projects is a constant source of frustration. 

• We call this gap the Innovation Orchestration Gap and, for manufacturers, bridging it requires wide-ranging changes with reverberations for the entire company.

• When companies do succeed in overcoming the various obstacles and closing this gap, they create connected, data-driven production environments, resulting in new opportunities for innovation.

The digitalization in the industrial sector gives leaders a cornucopia of opportunities for innovation. Some of these opportunities are so large that companies sometimes find it hard to see the wood for the trees. This blog post looks at how to ensure your production innovation and digitalisation projects hit the mark despite these challenges.

There are many pathways to digital transformation, but whichever pathway production managers embark on, they are inevitably destined to stumble over one major barrier to innovation – the gap between intended and actual outcomes. This gap between your ideas and innovation outlay, and your new, innovative processes and solutions is what we call the Innovation Orchestration Gap (IOG).

This blog post explores the challenges along the road to industrial digitalisation within production environments and illuminates the potential that can be unlocked when companies succeed in closing the Innovation Orchestration Gap.

Identifying hurdles to production innovation

Production processes are a major determinant of manufacturers’ profits. It is production processes that determine manufacturing output, adherence to agreed delivery dates, product quality, cost effectiveness, sustainability, etc.

Production environments are rarely created from scratch and many were not designed to achieve the outcomes expected of them in today’s era of digital transformation. Most production environments have evolved piecemeal over an extended period of time. As a result, the ability to meet key business goals is dictated by a fragmentary patchwork of systems and processes.

As an innovation service provider working on client projects, Zühlke encounters the same hurdles to innovation in production environments again and again. They can be summed up as follows:

• The brownfield problem: Production facilities tend to be populated by a mix of equipment commissioned from a variety of manufacturers over a period of decades. This fact combined with a silo mentality makes effectively implementing big bang data-driven innovations much more difficult. Realising connected manufacturing means overcoming proprietary protocols featuring minimal upstream and downstream data communication or even an almost total absence of accessible data.
• The gap between OT and IT: Data-driven production solutions require optimised data flows between source resources – usually managed by operational technology (OT) – and the IT infrastructure in which they are embedded. Manufacturers have lots of valuable data shut up in all sorts of different systems, with no means to retrieve it for reuse in other contexts.
• A lack of data analytics expertise and no single source of truth: Once the data extraction hurdle has been overcome, the next hurdle is often a shortage of analytics experts combined with the absence of a centralised data platform on which these experts can collaborate with process specialists to iteratively identify and solve problems.

If you want to explore the causes of the Innovation Orchestration Gap in production innovation in more depth, you might like to read our blog post “The digitalization obstacles hurting innovation outcomes”.
 

Overcoming obstacles and exploiting the full potential of a connected production

The bad news is that there’s no one-size-fits-all solution for closing the Innovation Orchestration Gap in your production. And the good news? The good news is that there are a number of approaches that can help manufacturers overcome the above obstacles to industrial digitalisation.

Create your source of truth

Enhancing efficiency by automating physical production processes has been standard practice in manufacturing for decades. A key point is that running in parallel to every physical process is an information process controlling the physical elements of manufacturing. It is common for this to give rise to high levels of highly localised optimisation, but often at the expense of the whole.

The problem is not a lack of data – production environments can produce terabytes of data per day. The challenge is to integrate this data, deliver comprehensive data connectivity, and use it to generate global added value. If you want to use data to leverage potential, establishing a reliable pool of production data is crucial. This can be used to standardise and automate information flows, ensuring that they are transparent, responsive and near real-time.
 

The first step on the road to connected production environments

If you were to build your manufacturing from scratch, you would build it so that every byte of data collected had a clear purpose – whether it’s simply measuring the status of a process, or whether it’s optimising availability, productivity and quality.

Sadly, in real life things tend to be a bit messier. Real life manufacturing involves decades-old equipment operated using retrofitted third party control units and run and maintained by a range of different departments. To avoid drowning in complexity, it is therefore important to build your pool of data piece by piece for each process or system sequentially. 

Start your connected production journey with a ‘data map’. This should start by recording your data sources (equipment control units, sensors, IT systems) and by identifying any essential information which is available in analogue form only.

The next step is to select a production unit you think offers significant potential and initiate a technology proof of concept. Machinery and sensor data from the selected production unit are collated in a data lab or on an analytics platform and combined with data from your ERP system, manufacturing execution system (MES), maintenance system, quality management system and input from staff. Using available data analytics tools, this pool of data is then analysed by your specialists together with data scientists. Once you have verified (using your proof of concept) that your plan for digitalising your production unit is sound, you can roll it out step by step to the company as a whole. At this point you can also develop a roadmap for converting any non-connected or analogue data to a usable form. The result is the systematic emergence of a fully digitalised, connected manufacturing environment.

Use data to boost the availability, performance and quality of bottleneck equipment:

Overall equipment effectiveness (OEE) is a standard for measuring manufacturing productivity. It covers three key factors:

• Availability: What is the total available run time? How much planned and unplanned downtime is there?
• Performance: How many products are manufactured? How efficient is the manufacturing process?
• Quality: What is the reject rate for the manufactured products?

OEE can vary widely depending on the exact type of manufacturing. In its projects, Zühlke has encountered OEE's ranging from 30% to 70%. Turn that around and it means that the scope for optimisation that we can aim to leverage ranges between 70% and 30%.

However, you can´t realize radical improvements with physical automation alone. It can be achieved by using a data-driven approach – though this too has its limits. 

In production environments, a data-driven approach is the key to overcoming the separation between organisational units, each with its own competencies, systems and ways of working. As long as teams operate in silos, information flows will always suffer from lag, inaccuracy and slowness. 
 

The first step on the road to better performance in manufacturing

Lots of companies perform only a very rudimentary measurement of OEE, which is anything but real time. Your first step should therefore focus on your bottleneck equipment, for which you should build a data-driven OEE, where possible using equipment and sensor data supplemented by input from staff. This information will be used both to identify areas offering potential for improvement and to gradually improve your pool of data.

One of my clients has a production process which uses a conveyor belt system. The system suffered from frequent blockages. The company had failed to recognise that these incidents were leading to downtimes of up to two shifts/16 hours, and that these downtimes brought the entire production process to a standstill. A data-driven solution was used to resolve the problem. The project team used equipment data to implement a solution which uses conveyor motor power consumption to detect reductions in OEE. Impending blockages are detected by a 2D camera-based machine vision system, which then stops the conveyor belt. The company is now able to detect potential blockages in advance and automatically reverse the conveyor belt for one meter to remove them.

Modernise and harmonise your IT/OT architecture

Production environments are created gradually in a piecemeal fashion. It’s difficult to maintain a consistent vision and integrated approach over decades of changing strategies and investments. 

We often encounter large, monolithic applications developed to perform specific business-critical tasks, but which fail to consider connectivity. The result is lots of data siloed in lots of different systems, which is difficult to extract for use in other contexts. In addition, we often come across proprietary systems and equipment with proprietary communication protocols.
We have to deal with a patchwork of legacy technologies, often from multiple vendors, typically with little connectivity between systems. In addition, for security reasons, these systems are often deliberately maintained separately. 

The ideal for a connected production environment is a common architecture covering and linking both OT and IT, with a shared (e.g. cloud-based) data/analytics platform forming the integration layer. Achieving this involves demolishing technological silos and breaking down barriers between organizational units and systems. One-to-one data connections are replaced by any-to-any communication. The ultimate goal is data-driven manufacturing. 
 

Radical production innovation extends far beyond production

A few minor tweaks here and there is never going to deliver radical production innovation. Working systematically to bridge the Innovation Orchestration Gap in your production, by contrast, can deliver huge potential. You can expand your focus to the production process as a whole or even the complete value chain. Such an approach enables you to target large-scale changes, with the potential to transform your entire business.  

For more fascinating insights into innovation and industrial digitalisation in production innovation, read our blog posts “Product Innovation means Business Transformation” and “Production innovation: the digitalization obstacles hurting innovation outcomes”