The firm's software, ProFicient, is a quality software platform that gives manufacturers a real-time view of manufacturing operations, allowing them to control quality at each point of the product lifecycle.
Clients in food and beverage include Ben and Jerry’s, Michael Foods and Nestlé Waters.
Martyn Gill, managing director, EMEA, said in the food and beverage industry, if you look at the industrial pyramid behind it all there are different layers in a business.
Case study: Ben and Jerry's
To track quantitative data, the ice cream manufacturer had previously been using a paper-based system. Operators would take individual readings and calculate an average to plot on a paper chart. Quality assurance personnel would then perform manual calculations to compute trends and create reports.
Ben & Jerry’s created run charts within ProFicient based on Six Sigma data from the plant floor to determine variations specific to each individual line, as each had a different run capability. Reports now run in seconds as opposed to the eight to 10 hours that quality assurance personnel previously spent calculating data.
Nina King, quality supervisor, Ben & Jerry’s, said: “By utilizing InfinityQS ProFicient to implement SPC and Six Sigma best practices across our manufacturing processes, Ben & Jerry’s will continue to identify opportunities for cost savings and ensure the highest level of customer satisfaction."
“At the very top they have SAP, Oracle and then the layer beneath that they have what they call Level 4 which is manufacturing intelligence or manufacturing execution systems (MES) which are software solutions and devices that can tell the process lines where the goods are being, at one end put on the line, at the other end packed and shipped to the consumer,” he told FoodQualityNews at the GFSI conference in Berlin.
“Along those lines there are devices, like sensors for motion and optical and PLCs, in-line checkweighers and all these devices have electronic outputs of one form or another. Our solution is able to connect to these third party devices and collate the data so that information is coming in automatically that allows the operator to make decisions.”
Monitoring beverages with BRIX
Gill gave an example of the move in the US, especially New York, to reduce sugar content in soft drink beverages.
“That resonates with the soft drink companies as they are thinking ‘we’re not going to sell anywhere near as much as this so we need to do something about this’ so that nutritional facts on the back of a label says so much sugar content and it’s not 17 teaspoons full of sugar in this bottle, there is only one or a substitute or an additive to sweeten it up.”
Gill said this has encouraged one company in particular to introduce adjusted processes to monitor the reduction in sugar content.
“This is done by a term called BRIX which can detect the composition of a soft drink so it might be carbon dioxide, the liquid or glucose content and there is a figure that comes out and it is called so many degrees of BRIX. That figure is what soft drink companies use to monitor the amount of sugar in a drink,” he said.
“There is one instrumentation manufacturer we work closely with that send out this data and it’s all automated so they can detect it while the bottle or the liquid is moving down the line, they can tell you what is in there but they also do it at a laboratory level.
“They have Laboratory Information Management Systems (LIMS) where this detection can go on with the syrups and whatever goes in before they place them on the line and often that is the best place to start, whereas our software can monitor in-line, the process activities, the quality, throughout and yield, ultimately, but LIMS can make sure it is right before you put it out there.”
The firm said regulation from the Gulf Cooperation Council (GCC) states saying they have to fill what they say on the bottle, instead of it previously being expected to go to the brim, will save drinks companies ‘millions’.
Real-time monitoring
Jason Chester, key account manager, said it is a variable process as there are so many factors.
Nestlé Waters case study
Nestlé Waters needed to standardize on one solution across all facilities to complement existing IT infrastructure. They were operating in both LAN and WAN environments and needed to maintain their IT framework. The firm had been using a paper-based system to collect and analyze data.
It now has real-time visibility over production processes—both within the individual sites and from the corporate level across 26 factories. The firm is using InfinityQS software to review sampling frequency optimization and in-line monitoring.
"There are many advantages to having a computerized system that sometimes, in the fog of having this powerful tool with its many capabilities, one can over look," said Julie Chapman, quality systems manager.
"It is ultimately easier for the operator. Even with the minimal computer skills many of the operators had in the beginning, the overwhelming consensus is that they prefer using InfinityQS over a paper system."
“You may have your recipe for that particular product but there are subtleties in the variability of the raw materials, altitude of the plant or the atmosphere can make a difference to how the sugar reacts with the mixing process,” he said.
“Monitoring that in real time throughout the production process can help the organisation keep that within tight parameters and the argument is if you control the process well enough, the quality will be ok at a LIMS perspective because if you get the raw materials and the mix right you will have a good quality product coming off the production line.
“It is capturing all that data and if you look at the amount of dollars that manufacturing enterprises have spent on sophisticated technologies around ERP, marketing, sales and look at the manufacturing process itself it is still in the dark ages in most organisations.
“They are still using pen and paper to record key quality and process parameters in what we would believe are quite sophisticated brands.”
Chester said it provides the enterprise visibility layer within the manufacturing process.
“You’ve got all that data coming out of PLCs and gauges and I would call that dark data as a lot of companies don’t leverage that, it is stuck in the PLCs, they don’t look at it and that is a lost opportunity.
“At the end of the day it boils down to three core characteristics cost, value and risk. Cost being producing what you need to produce at the optimal cost using the optimal amount of raw materials, energy, human resources etc, the value being represented in the quality of the product and the risk is how do you make a proactive manufacturing environment to be able to predict where risks might be.
“The final point is we would call it risk mitigation not minimising risk as you are always going to have risk, you cannot avoid it, so it is about mitigating that risk so when it does manifest itself how do you deal with it effectively.”
Traditionally inspections would be at the end of the production process, said Chester.
“So we’ll get these bottles coming off and be making visual inspections but whilst you are doing that product is whizzing past the line and if you’ve then got an out of spec situation, by the time the operators have responded to that you’ve got a lot of wasted products coming off the line.
“That impacts companies in many different ways, obviously from a pure cost perspective as they are wasting product but things like sustainability by wasting materials and energy throughout the production process. So having it at the end is, in my view, disastrous, what we need to do is do it throughout the process.”
He gave us examples of how data could help day to day business.
“So we are monitoring the fill levels as you might have a line that has 12 different fill nozzles, how is the performance comparing against those different fill nozzles is one particular causing an overfill or is another causing an under fill so that maintenance can focus in on correcting that particular fill head.
“Is one shift better at controlling the process than another shift, so does the other shift need better training or monitoring because they are not controlling the process as good or is one plant better than another.
“That soon creates a complex array of data…and a manufacturing intelligence solution like InfinityQS’ can monitor thousands of streams of information in real time and alert you to the areas where attention is required. If you illuminate that dark data without creating information fatigue then it can be valuable for an operator to know he has tight control of his process.”
Turning ‘dark’ data into something useful
Chester said it has what it calls a production rate data collector where it wouldn’t be valuable if every bottle coming off the line was measured as there would be too much data to create any meaningful value.
“So it’s about using sophisticated statistical analysis to be able to provide trends, correlations so the module analyses that real time data and makes statistical inferences from that particular period of time or volume,” he said.
“So it might talk about where the standard deviation is, where the outliers are, parts below, parts above etc so you can make informed decisions on a summary of the data when actually behind that there have been millions of data points going past.”
Gill said they monitor critical characteristics which could be net content control or it the wall thickness of a bottle.
“A good example is, look at the plastic packaging, if you go to an airport and pick up a bottle of water it’s squishier, the wall is thinner than it’s ever been,” he said.
“So now the thickness of the plastic has become critical to the process, they are saving money by doing that and by definition is it a critical characteristic and that is something they would monitor more frequently but it might not be critical to measure the thickness all over the bottle, whereas you might have 10 characteristics, one would suffice.”
Establishing normality comes back to what regulations allow, said Gill.
“You will have a tolerance or parameters around any characteristic like net weight, composition, the BRIX and as time goes on and you get better at what you do you can tighten their tolerances and become a better organisation and that makes them more competitive.”
Chester said when it does a trial it is about organisations changing their mentality and operators perceiving things differently.
“It can be used by organisations as a catalyst for change as what often does not happen is they’ll take the system and just duplicate what they were doing manually and automate it,” he said.
“They’ll use it as an opportunity to say ‘right we need to have more PLCs in the process so we can capture more data, we’ve changed the quality procedures, we’ve changed the terminology’.
“So we’ve had instances where we’ve gone in and done a pilot in one line and we’ve done that and the company has realised huge benefits just from that one pilot, on one line in two weeks and then the challenge is how do we replicate that in a global organisation and that can take time to deploy across their enterprises.”