Almost all the businesses and organisations I speak to know there are benefits to digitalisation, but for many the time, cost and uncertainty is too great to implement. There are pockets of digitalisation, and in some sectors like glass there are companies that are really far advanced in adopting digital monitoring and process optimisation. But in general, manufacturing is still very manual. And while most people are aware of artificial intelligence (AI) and digitalisation, the majority don’t think it’s immediately applicable to them.
The reasons for this are many, mainly around cost and time but also around skills and the fact that many pieces of equipment or plants have long-lifespans such as kilns, some of which have been around for close to 100 years. They have equipment that works perfectly fine and doesn’t need replacing, so why would they bring in a new digitalised version or fit expensive monitoring equipment to an end-of-life plant?
Commercially available solutions often involve expensive instrumentation and digital infrastructure to measure activity and processes in real-time – what industry really wants is affordable solutions and quick wins to work with what they already have.
What data exists?
The first question I always ask when I speak to a business is what challenges they’re facing. It might be trying to reduce levels of scrap or reduce the amount of chemicals or energy within their process.
The next is what data do you have? It doesn’t matter if this data is written or digital, I just want to know what they record. Then comes the challenge of working out if the data they have can help solve this issue – and nine times out of ten it can.
Often without realising it, industry is sat on vast amounts of valuable data that’s often costing them money to store. Industry collects data for continuous improvements and troubleshooting but often fails to exploit its full potential.
Then you have the inconsistency of data and how it’s recorded. Industry has highly experienced, skilled staff and operators, many who have worked in the sector for many years. Some will record regularly; others might forget, and some don’t write anything down. They’re so good at their job they don’t need to, they can tell by a sound in the vessel, a smell or just by feeling the end product that something isn’t right and will instinctively change things without the need to record it. Engineers have the data sets in their heads; they know what works and what doesn’t.
What the process?
That’s where partnerships with organisations like FISC and CPI can help. As scientists and engineers, we speak the same language as plant engineers, but we also have the industry knowledge and expertise to take data, combine it all together, digitalise, then use machine learning to generate possible answers to problems.
While industry has so many skilled people there is a huge skills issue around digitalisation. Businesses need to employ people in operational roles to keep production lines running. If staff are employed in purely digital roles, they are often focused on the day-to-day of the plant and don’t have the time to focus on more explorative work.
- For some businesses it’s a case of encouraging them to install a sensor here or there to get the data they need. Sensors have come down a lot in price; it’s about giving industry the confidence in the value and ultimately commercial value of buying, installing, maintaining and calibrating.
- Data collection is messy. The first step is organising everything. I’ve worked with businesses who have spent weeks tidying up spreadsheets and others who have handed it over and we’ve cleaned, sorted and analysed it for them. It’s a jigsaw. Working out the inconsistencies, what one person calls something in a column as compared to another, with data sets of 200 columns and many years’ worth of data. It’s often about combining digital data with human knowledge and getting it all in the right place.
- Then it comes to analysing. It’s not just about looking at the data, it’s about asking the data the right questions. When it comes to modelling, if the data you put in is rubbish in, you’ll get rubbish out.
What benefits can digitalisation bring?
- Immediate fault solving: Digitising data then analysing and modelling it can have instant and long-term impact and gains. I’ve helped businesses to use data to identify internal faults that would otherwise have been impossible to detect.
- Reduced energy use, scrap and wastage and increased process efficiencies: I worked with one business using a fuel-fired system that through data modelling, we were able to make suggestions on how to adapt their processes to reduce their carbon emissions; they said one small change we suggested has the potential to save around £10,000 annually on fuel costs.
- Predicting the future: Data can be a crystal ball of sorts and through modelling allows predictions to be made around degradation of material, lifespan of components etc. This means maintenance can be planned and shortened, reducing unexpected shutdowns.
- Provide new income streams: Like above, an SME I worked with was able to use its data to provide a predictive tool of commercial value to their own customers, they are now working on how to expand their business from purely commodities based to a serviced-based business model.
- Retaining and attracting a skilled workforce: Rather than industry being scared of digitalisation I’ve found the opposite to be true. Many embrace digital as a way of helping them in their role. In many traditional industries there’s a lot of lifting and shift work carried out by an older workforce. Introducing robotics and automation is helping businesses keep older workers closer to and beyond retirement and also makes the sector increasingly attractive to the future workforce who value better working patterns and conditions.
- Upskilling: Linked to the above, digitalisation allows employees to learn new skills and contribute to a business’s success. Early adopters are enjoying a commercial edge over competitors through the efficiencies and benefits of data.
For more articles like this, visit our Digital Transformation channel.
