Within the UK food and drink manufacturing sector, there’s a £14 billion growth opportunity waiting to be unlocked. The details were revealed in a new report: Future Factory: Supercharging digital innovation in food and drink manufacturing.
The report explains how the implementation of AI, automation, and digital technologies are key to seizing this untapped potential. Leveraged properly, they can lead to accelerated productivity gains throughout the sector.
The importance of AI has been further compounded by the Government’s AI Opportunities Action Plan unveiled in January. It outlines how AI can help to “turbocharge” growth and boost productivity.
The value of AI and machine learning is clear. Therefore, if we take the UK’s food and drink manufacturing sector as an example, how does AI and ML work? More importantly, what’s standing in the way of progress?
AI applications in manufacturing
Hidden inside the plant and machinery of every factory in the world there is a wealth of data. Once unlocked, this data can help to improve the overall equipment efficiency (OEE).
AI and machine learning, alongside deep domain expertise, are key to liberating and contextualising this data.
Half of the world’s top 12 food and beverage manufacturing companies – including names like Muller, Mars, ADM, Weetabix, Hovis and Diageo – are working with IntelliAM to harness the transformative power of their data.
We work by installing sensors that harvest millions of data points within a variety of supply chain components, the data is contextualised into a wide range of categories such as speed, pressure, product, flow and lubrication timing. This is then overlaid with reliability data indicating why faults occur.
These faults and problems can range from issues with vibration and oil condition to temperature of induction motors and loading of Programmable Logic Controllers (PLCs).
Once we have the knowledge of these factors, we equip the sensors with effective alarms, allowing for the health and efficiency of equipment to be monitored. This forms an individual stamp for each component that highlights crucial information such as finding the root causes for errors or mitigating future process shortfalls which, in turn, increases productivity.
For one of our clients, we implemented an OEE analysis and predictive maintenance system which harvests 400 million data points per month. This discovered consequential data that enabled us to predict future stoppages – through this non-invasive method we were able to increase their line performance by 6%.
Exploring the barriers to AI and ML adoption
At present, the top manufacturers are only accessing around 1% of their potential data.
For long enough, there have been hurdles in the industry which have limited production leaders from shifting their mindset be open to these new, transformative systems.
Yet while the Future Factory report states that 75% of the food and drink industry values the benefits of digital technologies, it also explores how they are held back by several cited obstacles.
These perceived barriers include the ability to instantly prove return on investment, negative preconceptions of AI and how to integrate it into legacy systems and equipment, as well as a significant skills gap, and rigid food and safety procedures.
But what if these perceived obstacles are more imagined than actual barriers? Mental roadblocks rather than real-world challenges?
Food and drink manufacturing is caught in a vicious cycle. Financial pressures restrict technology investment, leading to a stagnation in productivity, which, in turn, limits further capital investment.
But manufacturers don’t need to rebuild factories or invest in brand-new equipment. The answers lie within their existing assets.
Integrating AI and ML into the existing food production process
Machine learning that integrates with existing assets – no matter the make or age of the machine – means companies don’t need big capital investment to achieve the first steps to converge with advanced technology.
Another highly voiced concern in connection to AI is around job displacement. However, AI and ML work most effectively when they are coupled with domain expertise. A knowledgeable, well-trained workforce will always be needed in order to deliver impactful results.
AI and machine learning need teams of engineers to tag, code, and instruct the system so it can learn the algorithm to become self-sufficient. AI is therefore contributing to creating talented, skilled workforces.
It’s also important to address another misconception within food and drink manufacturing industry. Many believe that to get ahead of the curve and be a part of the AI and machine learning movement they need to abandon legacy systems and replace them with brand-new expensive machinery. This is a major misconception.
There are millions of data points hidden inside existing plant and machinery. They just need the right tools and technologies to liberate and, most importantly, contextualise them.
Having access to in-depth data insights helps to drive more informed decision-making, too. Manufacturers have the power of foresight – anticipating and fixing problems before they occur and determining training requirements.
Seizing the AI and ML opportunity
The challenges outlined in the report aren’t as difficult as they appear.
Data can be extracted from all machinery – regardless of the model, brand, or age.
Factory floors can continue business as usual whilst asset data is gathered in the background. This data can then be used to bridge productivity gaps and drive manufacturing forward.
This is more important than ever given that global food demand is always increasing to support population growth. Over the next 25 years, we’ll need to produce more food than humanity has ever produced before. This means food manufacturers will need to embrace technology and innovation to help meet demand.
Ultimately, whether manufacturers are ready or not, technology convergence is coming. AI and ML are redefining what’s possible in the food manufactuirng sector.
- Data & AI