“In a flour mill, with the data of the real-time sensors, you could adjust the pressure of the rollers or send the flour through more cycles to adjust the granulometry, or maybe decide to pass it (or not) through more plansifters or purifiers. The beauty of this is that you don’t have to wait until the product is ready to process again, or accept that some product will be lower quality because of the time involved in taking a sample, analyzing it, and then adjusting the machines. The quality loss is minimized.”

Prof. Gustavo Sosa Industrial Mechanical Engineer Licensed Grain Inspector MBA Project Management SOSA – Engineering Consultants ing.gustavo.sosa@gmail.com

Currently, digitalization and Internet of Things have became buzzwords that everyone throws around without know much about them. Seriously, it is starting to look like a business fad. In this article I will try to help you make sense of everything and save you from being scammed several thousands of dollars. Digitalization simply means putting sensors and actuators in everything, so you can control all the machines at distance, and also to provide real-time information on how they are working. A simple example is the speed switch in your bucket elevator. If the speed in the boot shaft decreases, it means the belt is slipping. An alarm should sound and everything should be stopped until the problem is solved.

Most of the value digitalization brings to grain processing is in the reduction of maintenance costs. Minimizing stoppages (by isolating them to a single machine), early detection of fires, monitoring bearing condition, monitoring belt condition, etc. Think about the downtime you had the last time an elevator got blocked, and it blocked all the other equipment before it. Or if you had a dryer that caught fire. Then you had the downtime while replacing it, the cost of removing the old one, the cost of the new one, the grain that got spoiled. That value is huge, but we are only scratching the surface of the possibilities.

I think the innovation in milling would happen in two different paths: 1) Going back in time with the technology and start producting whole grain flours that are actually whole grain, and not just bleached flour with a bit of bran. New mills, separators, and purifiers would have to be developed for that. Small scale, completely automated wheat mills would allow for local production of flour that is more nutritious, and appealing to the taste of millenials and their children. Most people would still consume standard white flour, but there is a huge niche market that I don’t see anyone trying to service.

2) Applying wet milling for the large scale production of gluten free flour. The future of mass consumption is mass customization. People should be able to choose the composition of a single bag of flour, starting with gluten content, but also vitamins, enhancers, bran, baking powder, preservers, etc. It actually is quite easy to develop a system that customizes like that, but the main raw materials have to be obtained through wet milling, which is not widespread.

Technically, milling is a process industry, but (unlike petroleum) there is very little technology available to continuously monitor the grain, flour or oil itself. In industrial control you have what is called a control loop. If you need an operator to setup the machine, take samples from time to time, analyze them in the lab, and then adjust the machine accordingly, you have an open loop. When the machine is able to sample and analyze the product by itself, and adjust itself accordingly, that is a closed loop. It is a closed loop even if the program that controls the process is in a server in Poland and not in the machine in Turkey.

There are some instruments that allow real-time measurements of flour quality. For example, the Perten NIR DA 7300, which provides moisture, protein and weight measurements, straight from a running pipe. Using the values obtained from the DA 7300 you could control the mixing of different grain or flour batches to maintain uniform characteristics. This is the one that I know, but I am sure there are some competitors manufacturing similar things. The sad part is that, even in large scale mills making many millions of dollars a year, you will find lab technicians running around with the samples and the results.

There are also optical sensors from Bühler that can check the color and the presence of specks in the flour, and others that can measure particle size.

For example, in a rice mill, you could know in real time the percentage of broken grains. If your hullers and your polishers had hydraulic or electromechanical regulation, you could adjust them instantly to decrease the percentage of broken kernels, and also adjust the graders downstream so they catch more of the defective grains.

In a flour mill, with the data of the real-time sensors, you could adjust the pressure of the rollers or send the flour through more cycles to adjust the granulometry, or maybe decide to pass it (or not) through more plansifters or purifiers. The beauty of this is that you don’t have to wait until the product is ready to process again, or accept that some product will be lower quality because of the time involved in taking a sample, analyzing it, and then adjusting the machines. The quality loss is minimized.

So you have the sensors on the machines themselves improving reliability, and the sensors that control the product improving quality. How do you handle all that? Remember it is not just the sensors, you also have quite a number of actuators; little instruments that DO things, not just SEE them.

The system that handles all this is a SCADA (Supervisory Control and Data Acquisition). It is not a specific software, it is the integration of sensors, actuators, PLC’s (the little computers that you see in control rooms), communication networks (cable, radio, or wifi), and HMI (those drawings of the machines where you can touch and see what it is doing).

PLC means Programmable Logic Controller. It is a small computer, many times without any screen, that is wired directly to the sensors and actuators of your machines. It can be programmed very flexible so it controls the actuators in response to the signals of the sensors.

Most major PLC manufacturers offer their own versions of SCADA. My advice is to choose one supplier (Siemens, ABB, Omron, whatever) and stick with it, using their software for everything. It simplifies installation and maintenance, and lowers your costs tremendously.

The communication network is a very important factor. Many people are falling in love with wifi and cloud computing, but I live in a Third World country and wouldn’t trust my facility to internet. Radio signals can be good if you have a very large facility, where the wiring would be expensive, but for the most part I prefer cable. Call me a caveman, I don’t mind. Radio is prone to problems with interference. Cable not so much. And the wifi may get down at any time.

You also have to choose the network of communication, how the machines and computers talk to each other. The most common protocols are Profibus, Modbus, Ethernet, and Profinet. The ruler now is Ethernet, because of performance and flexibility, but some other option may be better for you. Most likely, you choose Ethernet to provide the physical and data layers of the network, and then choose the application layer depending on what is common in your area, because you have to hire someone to maintain it. In Europe you would use Profinet. In South America, probably Modbus. See who sells more PLC’s in your area, and follow their advice. Most of the technicians in your country should have gone through their training, which is usually biased to a specific protocol.

Finally, the HMI (Human Machine Interface) is that nice program with drawings of your machines where you can click on anything and see real time information. The HMI is usually prepared by a third-party programmer based on the PLC software you use. Ask your trusted Industrial Electrician if he knows someone. I said Industrial, not a wire peeler. In large projects, they usually work with an Electrical Engineer specialized in Automation.

Personally, I would stay far away from wifi, internet-of-things, and cloud servers. Use cable, use your own servers, build a robust system. Out of Europe, the USA, or Canada, you will have a hard time finding reliable wifi, or even well trained technicians to keep your system running. It makes sense going back in time 10 or 20 years (in First World time) and apply that technology that is still innovation for us.

A good friend of mine (a CIO at a middle size company) told me his boss arrived one day at work, really excited, telling him about “the cloud”, how it was the future and why they should migrate everything to “the cloud”. Being polite, my friend tried to tell his boss that all the systems of the company already were on their own servers accessible from the internet. That is exactly what “the cloud” means. But his counterpart was a bit stubborn. The story ends with my friend renting very expensive space on the servers of an Internet company and selling the old servers, only so the boss could have his way. Don’t be stubborn, don’t get caught in fads, and you won’t lose money.