Make every drop of milk count

In recent years, the dairy industry has seen dramatic increases in raw material prices and energy costs. In parallel, many highly trained professionals have taken their hair nets off for the last time to head off on well-earned retirements. Despite the best efforts of HR departments, the industry is struggling to recruit young talent to fill their place, leaving significant gaps in expertise on the production floor. 

In response, it only makes perfect sense to strive for the lights out dairy approach by making the most of the latest advances in automated process control, not least, by exploiting the rapidly developing field of process analysis. As incoming milk accounts for around 50 % of production costs, this aspect of production is the obvious target for the use of process analysis in improving the outcome of the dairy process. Empowered by a constant flow of control data from the process, dairies can effectively get more from their investment in control systems by improving the use of key milk components in their products. At the same time, the dependency on manual labour and expertise on the production floor is reduced and the idea of ‘lights-out’ production takes a step closer to reality.

Figure 1: Visualization of process control in relation to the content of fat in milk as it flows into the dairy. The blue curve shows the original range of variation in fat. The red curve shows the reduced variation achieved with process control. This allows the production target (blue dotted line) for fat to be moved closer to the industry specification of 1.5 %. Significant gains in output can be made with no impact on quality and with no risk of overstepping the 1.5 % threshold.

Up and running in a few days

While the concept of continuous process analysis is simple to buy into, implementing still demands a practical solution that is accessible for busy dairy producers. Naturally, most production managers would love to just pop in a sensor at the appropriate spot and let the automation go to work. In practice though, achieving the level of precision indicated in figure 1, requires a more sophisticated instrument to feed the system with a consistent level of highly accurate data. Process analysis therefore requires more than just a sensor. That said, advances in analytical technology now make high-accuracy, low maintenance process analysis a reality with installation often taking just a few days to implement.

The technology behind this revolution is a form of infrared analysis commonly called fourier transform infrared (FTIR). FTIR technology has been used for the analysis of milk since the 1990’s and today has become the gold standard for highly precise and repeatable measurements that enable producers to push ever-closer to targets for profit and quality.

Infrared reveals profit opportunity

First used in the well-known FOSS MilkoScan™ solution which is operated by hand in a control room or laboratory environment, FTIR analysis was a great step forward. It was also clear that much more could be achieved by getting automatic measurements directly in production.

In 1999, a solution called FOSS ProcesScan™ was developed by FOSS. To meet the demands of the process environment the instrumentation was enclosed in a robust cabinet making it insensitive to vibration, heat and humidity. Samples of milk were then presented to the unit every few seconds via a sample interface connected to the flow of milk. It was a huge success, but like all great innovations, there is often room to improve on the first version.

The recently launched ProcesScan 2 does just that. It still might not be the plug-in sensor that production controllers might want, but a raft of innovative developments make concept more accessible than ever. 

More data with less work 

While working on exactly the same principle as the first version, the ProcesScan 2 has significantly reduced the time and expertise required to keep the instrument running exactly as it should. For instance, a manual procedure was previously required to compensate for possible impacts on analytical performance. This was typically done once or twice a month. Now the instrument does this automatically every few seconds as part of every measurement cycle. Manual work is avoided and performance is improved with a significantly reduced cost of ownership and daily work for the operators. 

Other features such as built in ID chips that continuously monitor conditions on vital instrument components also serve to make running as smooth and simple as possible. What is interesting of all for would-be process analysis users, is the fact that the FTIR based solution is super fast to implement. The FTIR technology works on fundamental information derived from the mid-infrared spectrum. This produces strong and sharp spectral peaks giving a very clear signal and making the whole process of installing and adjusting the instrument to local conditions a straightforward task.

Figure 2: Secure ‘one-way’ communication with control systems with the option to use secure cloud-based services to help run the instrument.  

Secure, non-cloud interface to control systems

The solution includes an in-built Linux Processor, so no extra PLC is needed to interface with a PLC or Scada system. The open platform communications (OPC) standard is used as it is one of the most important communication standards for industry 4.0. The OPC/UA PLC interface is a state-of-the-art interface that does not involve the use of cloud technology thereby promoting cyber security. However, there is still the option to use cloud-based services to monitor and maintain the performance of the ProcesScan 2 instrument. This is because the servers can only respond to input messages and can never initiate communication with clients. This paves the way for a range of cloud-based services from FOSS with top security (figure 2). In this case, no data from the local control system is shared with the analysis unit which is monitored and managed online via a separate system with no interference or connection to the local control system.

Key points when considering process analysis

Process analysis has considerable potential to improve the performance of automated control systems, but achieving gains demands an impeccable source of data for those systems to work with. A simple sensor is not up to the job. However, this does not mean that effective process analysis should require a lengthy and complex project. The development of FTIR solutions such as the ProcesScan 2 now make process analysis a practical choice for any dairy seeking to make more product from the same milk while also reducing dependency on human expertise in the plant.