The key issues keeping data-minded UK manufacturers up at night

Posted on 4 Jun 2020 by The Manufacturer

What is keeping the UK’s data-minded manufacturers up at night - and, more importantly, what are they doing about it? James Smith highlights some of the key conversations that set the agenda at this year's Industrial Data Summit Online.

Organising the Industrial Data Summit series of conferences over the past three years has provided me with a fascinating vantage point from which to observe the evolution of the ‘data conversation’ within  UK manufacturers.

The digital pivot

What were often quite niche and technical discussions in 2018 have broadened to involve stakeholders from across companies, helping data scientists and technologists to ‘learn the language’ of their organisations, and accelerating the sector’s shift to a data-driven mindset.

Thanks to the impact of Covid-19, this year’s Industrial Data Summit was forced to go online at short notice – but the change in format didn’t stop delegate numbers from continuing to grow by more than a third.

Clearly an event whose time has come. Industrial Data Summit Online was restructured around a single plenary session, supported by a series of break-out ‘Discussion Room’ sessions.

These Discussion Rooms each focused on a different piece of the puzzle, allowing delegates to choose which of the expert-led manufacturing conversations they joined.

All of this content has now been made available for on demand access here  – see bottom of this page for further details.

Industrial Data Summit Online 2020 Speakers

L to R: Steve Pavlosky, GE Digital; Bala Amavasai, Stanley Black & Decker; Dr Alison McGurk, AstraZeneca

From goals to enablement

The tone of the day was set by the event’s first keynote speaker, Steve Pavlosky, Senior Product Manager – Data at the Edge, at GE Digital, who framed the day’s conversations by challenging manufacturers to “derive more value from the data generated by their equipment, by establishing what your desired outcomes are.”

Most manufacturers are failing to get the value that they are entitled to from the data being generated, according to Pavlovsky.

“All this data that’s being created and not creating value for your business is no different from the materials ending up in the scrap bin. Less than 1% of data is ever used to improve decision-making,” he said.

Getting this data waste stream analysed and contextualised is key to driving optimal operational performance, and he went on to cite the example of Procter & Gamble which had collected Manufacturing Execution System (MES) data from multiple production sites, in order to compare performance, create best practice and raise productivity and quality at individual plants.

“Using MES transactional data tied to their quality and process data, P&G brings it all to a central repository where it can be analysed and then compare the operations of all of their plants making a similar product and get everyone up to the top level. This has created a huge opportunity to drive efficiency across the organisation.”

Digitalisation Analytics Data Digital Big Data ERP -Stock

Image courtesy of Shutterstock

Data and predictive maintenance

The next speaker was Bala Amavasai, Head of AI and Lead Architect with Stanley Black & Decker, who focused on how the 175 year-old company had gone about transforming itself into a digital organisation, and how it was using artificial intelligence to deliver products on time and on budget, reduce waste and improve quality control through audio and visual systems for QA.

The exponential growth of data has been driven by huge amounts of cheap computing power in the cloud and edge, but Amavasai explained that industrial data still tended to sit in silos; “For most manufacturers, the initial task needs to be connecting data together, putting it all into a single data lake, something that requires connectivity.”

Stanley Black & Decker has seen good results in predictive maintenance, by continuously monitoring equipment systems to allow the business to answer the question, ‘How much longer will this machine last after being in service for a given period?’

These results were only possible, according to Amavasai, through first engaging effectively within the organisation; “AI is just software, with a lot of research in the background, but it’s basically still a software project. If you want AI to deliver effective outcomes, you need to make sure you have all the requirements upfront.”

Digital Transformation Data Insight Decision-Making Technology AI Artificial Intelligence - Stock Image

Image courtesy of Shutterstock

Building smarter factories

The role of AI in enabling a smart factory strategy was the focus of Dr Alison McGurk, Head of Architecture, Data & BI, Operations IT with AstraZeneca.

The company has rolled-out a digital transformation strategy across its OT environment, looking at how it can scale-up the manufacture of new drugs, and deliver manufacturing process, QA and the supply chain improvements.

According to McGurk, digital innovation and operational excellence lie at the heart of AstraZeneca’s AI strategy, which has already fed into creating four reference proof of concepts: Predictive Maintenance, Yield Optimisation, Quality Deviations, and Visual Analytics.

“The first two of these are examples of predictive analytics,” explained McGurk. “In these cases it is important to define the scope of the proof of concept before you get started, and understand how you are going to prove that the results are actionable.”

Having the sensors in the right place in order to get the right data of the right quality, and then being able to analyse that data with the right algorithms, all determines your ability to get actionable insights.

Digital Revolution - Network cables closeup internet broadband fibre optic - image courtesy of Depositphotos.

Image courtesy of Depositphotos

Manufacturing’s interconnected future

Up next was Brendan Rawle, Director of Interconnection EMEA, at the world’s biggest data centres company Equinix.

He explained that the increase in enterprise-to-enterprise ‘interconnection’ was growing much faster than general internet connectivity – and that manufacturing as a sector was seeing a 57% compound annual growth rate in this measure.

“Interconnection is how you consume other people’s innovation, and a foundation for the collaboration that will be important within and between sectors,” said Rawle.

Taking the automotive sector as an example, he said that the industry was increasingly interacting with adjacent sectors such as insurance, fleet management, in-vehicle entertainment, ride sharing.

As you move from an automotive sector to a smart vehicle sector, it’s going to rely very heavily on the interconnections within the ecosystem,” he continued. “Those that invest will have the advantage.”

CCAV and Meridian Mobility has up to £30m for projects that help make the UK the most effective ecosystem for self-driving technologies - image courtesy of Depositphotos.

Image courtesy of Depositphotos

Delivering your data strategy

The Manufacturer’s Jonny Williamson hosted a panel discussion featuring three leading practitioners from genuine industrial champions: Gerard Bartley, Global Master Data Manager, Jacob Douwe Egberts and the conference Chairman, Suresh Daniel, Data & Architecture Integration Director, Coats, and Keith Roberts, Global Parts Operation Data & Reporting Manager, Jaguar Land Rover.

Focusing on the role of organisational culture in enabling a data strategy to be successful, we heard the approaches taken by three very different businesses.

We learned that Jacob Douwe Egberts was a large and successful heritage business specialising in making excellent coffee, but with a more conservative approach to data – one which required more of an effort to create a business case.

In the case of Coats, efforts to embed a company-wide culture were made more complex by virtue of the broad range of customers from automotive OEMs to hygiene – each of these segments having distinct personas, often with very different methodologies and cultures of their own.

This really placed a premium on effective leadership to sustain and develop a data-driven culture within the company.

JLR’s situation was very different again, as we took a closer look at their aftermarket business – one which had inherited a lot of systems from past mergers and acquisitions, and which made integration and standards the key priorities of the team.

With ‘line of business’ managers, across each of the manufacturers, working more closely with their data science and analytics colleagues, the panellists noted that there was a ‘common language’ developing that enabled industrial data resources to be leveraged for greater impact.

This shared vocabulary is perhaps one of the most tangible results of the past few years of running this conference – and one that may well create the greatest long-term value for manufacturers.

On Demand

Access more than eight hours of content on demand in order to improve your understanding of how manufacturers are exploiting their data to drive operational improvements.

In addition to the plenary session content described in this article, Industrial Data Summit Online featured eight ‘Discussion Rooms’ each led by experts, and focused on the following key issues:

Data Collection Strategy | Eliminating Data Silos | Data Quality & Integrity | IOT & Sensors | Data Integration | Monetising Data | Predictive Maintenance | Business Intelligence | AI & Machine Learning

Upcoming Summits from The Manufacturer – click the banners for more information and to register: