Drowning in Data

Abstract

In their conversation, Mariella Barra and Alistair Williams delve into the critical role of data-based decision-making in the food and drink manufacturing industry, as well as the implications of Industry 5.0 on supply chain traceability.

When discussing supply chain and traceability, it is inevitable to talk about data. As food professionals, we are used to collecting and analysing data - essential to fulfil our due diligence obligations. We record raw materials delivery and batch codes, and we must keep meticulous record on whether we have achieved a process’ critical control points. What else we do with this data once they are collected, however, is an entirely different story. Often, we are satisfied with meeting our legal and ethical obligations and fail to go further, potentially missing out on valuable information that could lead to better business decisions. At a recent training session I heard, for the first time, that ’in 5 years there will be two types of food and drink manufacturing companies. One type will have embraced data-based decision making and the other will have been acquired by companies that have embraced data-based decision making’. Like most sweeping statements, it is necessary to peel the layers of this particular onion.

Industry 5.0 is already in the ascendancy, predicated on the ability of all manufacturers to leverage machine learning and artificial intelligence to analyse big data sets and produce reliable information. This enables timely, affordable, and secure investments or interventions. The jargon and complex concepts surrounding Industry 5.0 often make it seem disconnected from the practicalities of real-world food and beverage production. With the vast majority of the producers being small to medium sized enterprises, the current economic maelstrom has stretched all resources to the limit. Time, money and appropriately skilled people are all in short supply.

With the constant increase in audits, compliance requirements, and reporting obligations, it can often feel like managing paperwork and records has become the primary focus, while making the products has become secondary. The sheer pace of change is staggering, and at times feels overwhelming. In reality, the technology to collect, extract and analyse data is becoming more accessible by the month.

With affordable, secure solutions enabling sensors to be retrofitted, cloud-based apps that use low code or no code tools to collect, aggregate and accessibly store vast amounts of data with an air gap between business-critical infrastructure, it is only the imagination that limits the first steps towards a quantum leap in business intelligence.

Mariella Barra is the chair of the IFST Food Processing Special Interest Group and Alistair Williams is the F&D Theme Lead at NMIS. It was during a conversation about the F&D supply chain data environment that they focused on the Q&A that makes up this article.

With their different experience and common passion, it developed into a wider ranging discussion about traceability and data-based decision making in the industry.

This is the very core of the challenge. There is a tsunami of data flooding into our businesses, all of it important and some of it essential to ensure safety of the product and the teams. The ability to collect, identify, categorise and act on individual elements of the data stream will require machine learning and intelligent systems reporting by exception and flagging essential or critical events in real time.

This is the number one obstacle to a smooth transition to a digital factory. The amount of information essential to the production of safe food & drink can feel overwhelming; it is easy to forget when wrestling alligators, that you got in the swamp to drain it. Often, we celebrate our internal ‘fire fighters’ - those members of our team that can be deployed to rapidly fix a problem. Is it reasonable to suggest that the quiet, meticulous team members that stop problems arising can be easily overlooked? Data may disappoint but it doesn’t lie, so how do we find the common denominator? Current technology is affordable, secure and accessible. Single data point recording is essential. Whether it's a spreadsheet in the company cloud, a form recording real-time information with automatic timestamps, or the scanning of a QR code triggering an event and logging the record, acquiring accurate, timestamped, and accessible information is now a low-overhead activity, making backups a thing of the past. It is not possible to make the transition to Industry 5.0 without digital data, accurate and complete, timestamped and automatically and securely backed up.

Today this can be started with minimal investment. With the right mindset, Google and shared learning it is possible to start this journey now and start enjoying the benefits tomorrow.

Writing is thought to have begun around 3300 BCE, no wonder it is proving such a hard habit to break. There is absolutely no need for metal detectable pens and thousands of kilos of paper in a 21^st century manufacturing facility.

This is a big statement, and now we can put our money where our mouth is, literally. The point about foreign body contamination is somewhat of a reductive argument. The reason pens and calculators are blue metal detectable is to mitigate the hard plastic risk. Blue plasters however are used in high-risk food preparation regardless of the presence of a metal detector as we know it's impossible to eliminate risks. So, the use of highly ruggedised handheld devices is not new in the industry. Advanced in toughened glass screens and the additions of plastic screen guards when combined with adequate inspection and training will control the risk to the same extent as using a plaster or pen and paper.

With effortless, no code design, it is possible to create digital replacements for all of the paper systems. The inputs can be timestamped, making retrospective changes impossible and it is important to remember that picking one's name of a drop-down menu is as secure as writing one's name on a form. The digital version will be readable, securely archived automatically, and most importantly be accessible in seconds during an audit.

Collecting data is meaningless without understanding. Without accurate interpretation and knowledge, it is easy to be precisely wrong rather than approximately right. The Co-lab initiative at the National Manufacturing Institute Scotland, is designed for exactly this purpose. Professor Andrew Shirlock realised that there was a growing disconnect between the rapidly developing academic approach to big data management and the ability of industries, not just in food and drink, to access these game changing advances.

The confidence to engage with modern data acquisition and analysis is, without doubt, the biggest stumbling block faced by SMEs. Retrofit or integrated sensors can provide insightful and enlightening data. Cloud based apps keep the airgap between business-critical and engineering improvement data giving affordable and reliable security.

Low code / no code applications enable fast and accurate reporting by exception and bundled office software is putting advanced statistical packages withing useful reach of anyone that can write a spreadsheet.

With the democratisation of technology, real world, robust and easily implemented solutions are affordable and available for SMEs.

For example, considering Microsoft Forms. Included in the MS 365 suite of products Microsoft Forms is the ideal introduction to real time, secure data capture, back up and basic analytic and data presentation.

MS Forms is intuitive to use, a no code solution, which effectively draws and chooses question formats, and the Form automatically generates a QR code that pulls the form up on any devices configured to read QR Codes, independent of location and network.

Using individuals’ handheld devices, including regular mobiles, data is instantaneously captured, stored in the secure company cloud and available in either infographic, or Excel format.

With the move from Capex to Opex data capture is very straightforward to automate. The value this adds to a business, is that it ensures no gaps in the data set, allowing colleagues to spend their time proactively, delivering paybacks often measured in weeks and months rather than years.

Sensorisation and the integration of the Internet of Things (IoT) is a solution to missing data points. Knowing that production, training and maintenance record automatically complete, correct and backed up, gives Food and Drink manufacturing Professionals the confidence and peace of mind that data is 100% audit ready 100% of the time.

Freeing this group of professionals from menial and repetitive tasks enables them to focus on productivity and innovation. With increasing demands on the skills and time of food technologists and engineers, the ability to concentrate their efforts on continuous improvement, exploring alternative fuels, waste reduction and reporting etc is essential to remain competitive in today's global market. Making data work for us is a journey, and each individual company will have their starting place on the road. The simplest starting place is digitalisation – this, as remarked above, can be done with inexpensive tools – the first simple step which enables the rest of the journey is centralising our data and making them available in an electronic format. This enables us to move to more sophisticated approaches, using machine learning and a variety of tools for in-depth data analysis and correlation.

Especially for small businesses it might take some time and effort to implement this, but the advantage is to be able to make proactive decisions driven by data rather than take reactive actions – we can spot problems in our process or supply chain before their business impact becomes too severe, we can more easily implement cost-saving or rationalisation exercise, we can keep track of how our suppliers and providers perform, and we can keep our internal and external processes streamlined and coherent.

The National Manufacturing Institute of Scotland (NMIS) recognised the food & drink sector as a nationally significant asset. NMIS is part of the national High Value Manufacturing Catapult, an initiative to facilitate the journey to manufacturing 5.0 in the F&D ecosystem.

The national HVMC is a noncommercial collection of specialist centres (Catapults) and NMIS is part of Strathclyde University, and its purpose is to support and de-risk innovation and continuous improvement.

Working in areas such as augmented reality to assist in machine break fix and maintenance, remaking and repairing obsolescent damaged machine parts and the design and implantation of automation and robotic.