HomeMachinery & Lab EquipmentThe future of feed production Exchanging machine stops for robots

The future of feed production Exchanging machine stops for robots

by Andrew Wilkinson, MIlling and Grain magazine.

 

By 2050, with the global population expected to reach 9.8 billion, our food supplies will be under far greater stress. Demand will be 60 percent higher than it is today; so going forward the pressure is on our industry to adapt and grow in order to meet this likely future demand.

In the latest in a series of Famsun-led webinars, this edition titled 'Feed Production – Management & Efficiency', set out with the intention of offering potential solutions for bridging this gap.

The webinar begins with moderator Roger Gilbert, Publisher of Milling and Grain magazine and Jeroen Ju, Famsun's Marketing and Planning Manager, welcoming all of the one hundred or so delegates and thanking them for their attendance. Following the introduction, the first of two speakers to address the audience was Kim Jensen Møller, R&D Director of the Famsun Denmark R&D Centre.

Mr Møller has been working in the IT industry for more than 25 years, including 12 years within the agricultural industry. Throughout this time, Mr Møller has specialised in automation and management systems for the farming and feed industry.

Currently serving as the lead architect for the development of Intelligent Feed and Food Factories within Famsun, including data analysis for intelligent machine control, as well as the development of the next generation of Manufacturing Execution Systems (MES). Mr Møller also has a Master of Science in Computer Science and Economics, all of which places him in a great position to talk about the future of feed manufacturing.

Mr Møller begins his address by discussing the digital feed factory, paying particular attention to the digital tools typically contained within it. He also provides some background about why we need digital tools, what the benefits could be and what could potentially be achieved with their use at the feed factory.

In the words of Mr Møller, Famsun are a provider of integrated solutions from the farm to the table. To achieve this, it has divided this topic into a selection of different sub-topics or business areas, with each one specialising in one of the relevant areas. The business areas include feed milling, animal farming, grain storage, oilseed processing and modern agriculture. The last area is the business section that this webinar focuses on, with particular attention paid to the automation of factories and the intelligence systems that enable this.

Industrial evolution not revolution

The first business area that Mr Møller examines covers the various pieces of equipment that are employed in the typical feed factory. Before going into real detail on the subject, he starts by first examining what he likes to call 'industrial evolution', which breaks the 250 years of the industrial revolution into four different 'evolutions'.

The first 200 years saw the use of water, then electricity brought in to ease production. Then, some 50 years ago we had Industry 3.0 which was the third industrial revolution. This phase saw the addition of automation and computers into the production process.

And then today, we have what Mr Møller calls Industry 4.0 which is the fourth industrial revolution, where we have machines and computers able to communicate with each other, gathering data via sensors and then sharing that information with other stages of the production process.

According to Mr Møller, we currently have systems that are able to talk, so for this reason the production process will at least appear to possess the capability to make decisions on their own because they are Artificial Intelligence (AI) systems.

So if we want to introduce this technology to the feed production industry, according to Mr Møller, we need to first establish what our priority should be, in order to optimise our production and ultimately, our business.

Establishing what this first priority is presents a difficult process. He suggests we should first ask ourselves if what we need to add is more robots or increased automation?

Mr Møller argues that first and foremost we should be looking to get more knowledge and insight into our production processes. He continues that before looking to invest in more equipment, we should instead prioritise 'getting more knowledge'. Then by collecting data we can create a digital system as a foundation in order to support future developments, such as automation.

'By collecting this data, we will be able to yield a greater depth of insight into the nature of our production processes,' he says.

Mr Møller believes that we all know where we should start, we will also find out which tools are essential and if they are going to improve our businesses. This type of challenge does not only exist in feed manufacturing he argues, but they are found in any type of manufacturing, or any other area of business for that matter. Once we have collected all this data, only then states should we even think about adding more robots!

The impact of machine stops

One of the first areas we need to look into are machine stops.

According to Mr Møller, we all have machines that are stopping for some reason but we may not always know why they are doing so. We also need to know what their impact is on the business.

'All this information is required to actually know what impact that this has on our plant's performance, with this being one element of business that we could all do with being more knowledgeable about,' he adds.

Sometimes it's better to actually do maintenance before you have to, he says.

'We all know that we should send our car for service before it breaks down. Maybe we should also do that with our processing equipment now that we know how lomg it has been running and when we should do maintenance. We need to do this so we don't have breakdowns.'

While preventive maintenance as a minimum, Mr Møller argues that we should also look at employing a predictive maintenance philosophy, which is more complicated.

To achieve this effectively, we need to calculate how much the machine actually runs on a daily, weekly or monthly basis, as a system outage can actually allow you to carry out this maintenance when the system is not in use. A predictive maintenance philosophy can also help us to make sure that we have the right spare parts, as well as access to an adequately trained team who are able to carry out the maintenance.

We all need visibility in our manufacturing, as we need to be able to show the performance of the factory in order to know when it is working efficiently.

In fact, Mr Møller also believes that it is a good idea to share the factory performance schedule with staff so they are aware of what level and capability the machines and production lines can achieve. By just acquiring knowledge about how we are actually performing could increase the performance of the machines, he adds.

Enhanced information adds further value

The basic foundation here is machine flow, that automation can add to production through from robots on the production floor. With the addition of enhanced information technology, we add further value. On top of this, we are also getting real time data from all the production machines, states Mr Møller.

We have orders coming from customers, which are then used to directly control the machine's output. We need feedback from the machine, we need to mention our data from the production flow from the machines, we need to get notice of unforeseen events such as any alarms and we collect all this information, put it into a database and then we provide the feedback into our business systems about the performance. More detailed factors such as individual performance indicators and traceability, can be stored for use in future business system decisions.

The Supervisory Control and Data Acquisition (SCADA) system takes care of all these technical details, such as how to start, when to start, which machines to use, which recipe to use, as well as a variety of technical parameters. These technical parameters include the chosen equipment and the consideration of any possible production line issues such as bottlenecks.

The system then sends information back, producing data about the status of the production and processes. When it has collected all the data, the system will then send information regarding the condition of the product and where it is in the process back to the tracking system. In order to optimise production so that the best results for the business are achieved, the digital service platform will need to be redesigned.

Software architecture is one of the interfaces of communication between the components, states Mr Møller. We need a highly efficient network running 24 hours a day and we need to be able to have a flexible system where we can add new functionality or adjust the set up to accommodate customer requirements.

So, you can get information from manufacturers at the same time. Having this platform on your factory and the regular status update system, you can connect to all the systems running within the factory and we can communicate directly with the hardware by collecting data from the hardware or their sensors.

With a system like this, we can integrate all of the machines together. This is a horizontal integration and then we have a vertical integration between the production and the earpiece system. So, we are connecting our production to the business system, states Mr Møller.

Some of the user interfaces includes a warehouse dashboard, which displays key information about your running orders in your warehouse. The warehouse orders are going to be coming in momentarily and getting from the warehouse into production. Then the finished material will be leaving the production line, before heading out to the customers, with the system also able to display the shelf life of the different materials.

Improved traceability and efficiency

This system has the ability to tell us which raw material is used for a specific batch. This in turn allows operators to trace their raw materials back to a specific supplier; with them then able to produce a material report that tells them which products contain that particular material.

In order to ensure that traceability is maintained right from raw material to the final product Famsun have two Key Performance Indicators (KPI). According to Mr Møller these include all equipment efficiency and yield efficiency, with the former affording the opportunity to calculate the effective run time of machines, which in turn predicts the processing time of the raw materials.

This allows operators the ability to communicate with farmers automatically, as well for some Famsun systems such as the smart cost systems, the trucks and the smart card system, ensuring synchronicity in the manufacturing process from field to fork.

This also applies to yield efficiency, as from this data we can calculate how much product the production line actually produces, based on how much raw material that we actually put into the product.

Some intelligent machines add control to these different machines, as they're taking data readings from the machines and then analysing this data. Then based on this data and from the analysis of a lot of machines and a lot of production, you can customise how the machine is actually running in order to improve the performance and reduce the energy consumption.

Employing AI for ease and consistency

One example of a piece of equipment that employs such a system is an intelligent dryer. By adding a lot of sensors to this dryer, it is then able to communicate with the extruder.

The goal here is, instead of having a vast array of different controls, you need to only adjust the fan speed, belt speed and heat; all of which can be controlled directly at the start of your process, so you can achieve consistent moisture of your product.

So instead of having lots of different controls, that you have to do more or less manually, we have a much more sleek and intelligent control. So in this instance, less really is more.

Regardless of the product that is being produced, this system uses sensors to collect information, which can in turn ensure that the product produced is of a predetermined quality and remains consistent.

By using predictive analytics and machine learning AI, the performance of the dryer can actually be improved by avoiding over drying, thus resulting in a more uniform level of moisture. This should in turn secure a more stable level of moisture, whilst also seeking to reduce energy consumption.

NEXT POSTS

Subscribe to our newsletter to get the latest news from industry

Newsletter Registration

DOWNLOAD OUR APP