What does Big Data mean for manufacturing?

The notion of an Industrial Internet of Things supported by Big Data is gaining traction globally – emphasising the idea of consistent digitisation and collecting data from different sensors and systems across productive units across the industry.

The utilisation of that data, combined with other data sources, will bring unprecedented opportunities with it, but also new risks to business and society. Furthermore, it will define the modern manufacturing landscape and lead to better-integrated supply chains, interconnected systems and stronger co-ordination.

However, using Big Data in manufacturing is still in its early stages. Many important questions remain, for example, ‘how it will impact existing industries, value chains, business models and workforces’ and ‘what actions business and government leaders need to take now to ensure long-term success?’

Recent studies predict that manufacturing operations will become more geographically distributed in the future, which is driven by an increasing need for mass customisation and more sustainable production. Such a transformation will be supported by technological advancements like additive manufacturing, advanced robotics, the Internet of Things and Big Data itself.

In that context, the UK research council (EPSRC) actually refers to a redistribution of manufacturing, coining it as “technology, systems and strategies that change the economics and organisation of manufacturing, particularly with regard to location and scale.”

An additional definition illustrates redistributed manufacturing as a “connected, localised and inclusive model of production and consumption that is driven by the exponential growth and embedded value of Big Data.”

These definitions show that there are several dimensions of redistribution (connected, localised and inclusive), and that the value of Big Data arguably influences all of them. All three can be applied to trailer and component manufacturing as well, with some multi-nationals already investing in additional research on the topic.

The inclusive dimension, which is aptly described as an “inclusive” model of production and consumption, can be interpreted as a functional redistribution. This dimension is comparable to concepts of co-creation or co-production. Co-production refers to a “participation in the creation of the core offering itself”, while co-creation represents a concept that includes the idea that “value can only be created with and determined by the user in the consumption process or through use”. Value co-creation can occur with or without co-production, however both concepts illustrate that the end-user is part of the value creation process. To this end, the term ‘redistribution’ means a higher involvement of the consumer in the process of design or production.

Another dimension of redistributed manufacturing is described in the “localised” model of production. This emphasises a change in the location and geographical configuration of production facilities, with a particular focus on production scale and distance from production to a customer. The redistributed concept in this context implies a shift to a smaller-scale and more localised production.

At the Cambridge Service Alliance, we recently explored the impact of Big Data on the redistribution of manufacturing within the fast moving consumer goods (FMCG) sector. Our research revealed that the FMCG industry is focusing on Big Data applications – especially on external data sources – with the aim to better engage with the customer and to understand customer preferences, effectively making the design and marketing departments the main data users at this point in time.

One example to illustrate the power of Big Data in the context of market insights came from global dairy giant Danone, which uses analytics to compare a variety of sales data, including those from the competition. By analysing it, Danone was able to identify increasing sales of a specific Greek Yoghurt brand in the US – enabling it to produce and deliver the right product for the right shopper at the right time. With predictive analytics, Danone increased its forecast accuracy from 70 to 98 per cent, using data from a two-year history of purchases that included so-called ‘seasonalities’ as well as additional data from trends and promotions, combined with sophisticated algorithms to project forward.

Focusing more on the manufacturing side, another case of intense Big Data use is Coca-Cola. The global powerhouse has developed an algorithm called Black Book, which ensures that the consumer gets orange juice with a consistent taste 12 months a year, even though the main growing season of oranges only lasts for a short three-month period. The algorithm helps Coca-Cola find the right mix of ingredients based on an analysis of up to one quintillion decision variables and diverse data inputs – from orange sweetness and consumer preference through to weather patterns around the globe.

All businesses we examined as part of our research already used Big Data intensively – especially those that mainly compete over price – in order to differentiate themselves from competitors by understanding their customers better and engaging with them.

An interesting example of how that can ultimately lead to the redistribution of manufacturing is the story of Spanish fashion outfit Zara. While clothing businesses are traditionally motivated by the incentive to cut costs and therefore produce in low-cost countries like China or Bangladesh, Zara has been shifting part of its production back to low-cost countries in Europe recently. The move was not necessarily motivated by cost savings, but more so by shorter time-to-market, local expertise and closeness between design and production.

Reducing the time-to-market – at its core a logistics challenge – can normally only be achieved by ordering smaller batches, which is not feasible for many a high volume business. In the case of Zara, however, the focus shifted to fast small quantity batches sourced from Spain, Portugal and Morocco. While such a model may cost more, it shortens the supply chain and enables the company to react quickly to trends identified by using Big Data.

The Zara case also showed us that geographical closeness to the market is only one factor to be able to respond quickly. Internal structures are important as well – they need to be streamlined for the so-called ‘fast-fashion’ model.

Zara, for example, taps into its point-of-sales data to enable a market-responsive supply chain. To do so, it has developed a complex feedback system that is mainly fed through retail stores and social media channels. To ensure direct real-time access to all customer-related data, the company actually owns most of its supply chain, including retail stores, showing just how important integration and communication have become for modern manufacturing.

Dr Mohamed Zaki

For today’s manufacturing businesses – including those in the heavy vehicle sphere – Zara is a prime example for how effective data utilisation can lead to added competitiveness. It may not inspire a complete redistribution of manufacturing given the size and scope of a modern semi-trailer facility, but still has the potential to help streamline supply chains and make markets movements more predictable.

If manufacturers can’t innovate when the opportunity arises, they may lose their competitive advantage and be left struggling to ‘catch up’ with their competitors and, ultimately, market share and the accompanying revenue.

ABOUT: Dr Mohamed Zaki (pictured right) is leading the ‘Big Data and Analytics For Service’ research at the Cambridge Service Alliance, University of Cambridge. His key research interests lie in the field of Big Data analytics implications for service science, business model innovation and customer experience. His research uses an interdisciplinary approach of Big Data technology to address a range of real organisational problems to make better business decisions within complex service network.

Leave a Reply

Send this to a friend