A plug's game

By John Dunn

- Last updated on GMT

Related tags Automation

A plug's game
The two big challenges for food manufacturers today are flexibility and efficiency. One minute Tesco wants cartons of freshly squeezed orange juice. Two hours later Morrisons is ringing up for blackcurrant drink. How do you reprogram the production, filling and packaging lines to give you maximum flexibility yet keep changeover times to a minimum?

Most food firms know they have to maximise efficiency to get more out of their machines. But what are they doing about it? Many, it seems, still send managers wandering onto the shopfloor with pen and paper to ask the operators how many minutes of production they lost and why.

Simon Ellam is business manager for process automation at Siemens Industry Automation. He says the big challenge for food manufacturing is the need for flexibility. "Some customers have very specific batch requirements because they need to produce an orange cordial in the morning and a blackcurrant one in the afternoon. So on the primary side they have got to have flexibility. Then the secondary side, packaging, has to be flexible, too, in order to be able pack orange drinks and blackcurrant drinks. There's a huge amount of pressure."

Paul Everington, md of Tascomp, developer of industrial automation software, sees the need to improve efficiency as the big challenge for the industry. "Increasingly we are seeing firms looking to improve their efficiency, in particular targeting the utilisation of existing equipment, their overall equipment effectiveness. It's been around in other sectors for years but is just beginning to find its way into the food industry."

But perhaps the biggest challenge of all is integration, getting all that kit on the factory floor the modern, the old, the automated, the manual and electronically linking it under one smooth control system. It is integration of control systems that allows the production director to sit in his office and find out all he needs to know from the computer on his desk. He can see which machines have made which products and how many and when. Instead of reports on bits of paper he can see at a glance which machines have stopped and why, and how efficient the last shift was.

Only integration gives food manufacturers the flexibility that is being demanded of them, says Ellam. "But the problem is that food companies get all their fantastic, high-speed machines from different manufacturers and then they have a hell of a challenge on their hands integrating everything. If they have got to change a production line from Morrisons to Tesco in a matter of hours then it's a huge challenge if they have a load of disparate systems from a load of different machine manufacturers.

"There have been advances towards integration but ultimately the OEM [original equipment manufacturer] is king. And food manufacturers, unless they have got a very hard-nosed engineer to tell them what they need, get dictated to by the machine builder. A firm like Siemens will go to the end-customer and show them the flexibility that integration can bring linking the different islands of automation under one cohesive control system. And they love that, they think it's absolutely fantastic. But telling that to the machine builders and the plant suppliers is the hard bit."

Food manufacturing is a hybrid industry, says Ellam, which brings problems when integrating control systems. The primary processes deal with mixtures of liquids, powders, raw materials, and pipes, tanks and vessels. And that traditionally is a world that has always been controlled by distributed control systems (DCS) or process control systems. "Very integrated with a single operator view of everything that is going on."

On the other hand, the secondary side, the packaging end, is typically islands of automation, he says, with machines all doing their own thing. "You could have 20 machines from 20 OEMs around the world with 20 different PLCs [programmable logic controllers]. It's a huge challenge, 20 machines all with their own little operator control panel, all with their own communication and networking standards, and all of which are, potentially, going to be very difficult to integrate together."

If a machine will mechanically stand it, then electronically integrating it into one control system is well within the realms of specialist systems integrators, says Ellam.

An integrated control system can save a lot of manpower and time. "You can have one team in a control room who can see everything that is going on, right the way from raw materials to the secondary process. You have a common alarm systems, common fault logging throughout the whole plant, and common tools."

Fortunately there are now some food firms that understand the need to integrate control systems, he says. "They are quite prepared to take some pain as far as costs go because they know ultimately that the capital cost of a plant pales into insignificance compared with the operational costs of the plant over 20 years." He cites one customer, a chocolate manufacturer, which has gone for an integrated control system on an existing process. "It was very much PLCs and islands of automation. We put DCS in and straightway they were able to drop a shift, literally only a few months after commissioning."

The food industry has been quite slow at hooking onto the fact that looking at the efficiency of its equipment is a good starting point for gaining improvements in performance, suggests Everington at Tascomp. "When you are dealing with small food manufacturing companies you half expect them to be a bit behind the times. But you would imagine the larger companies would be quite switched on."

Review efficiencies

The problem, he says, is that many firms know they need to monitor availability, performance, and quality. They know they need to analyse downtime and to be able to understand the causes of their lack of performance. "But a lot of them attack the problem of getting that information by just asking operators to fill in how many minutes have been lost and for what reason. Even those firms that think they are more advanced are still putting computers on the shopfloor and asking staff to fill in spreadsheets."

However, a number of the larger food firms are now applying these techniques successfully, he says, probably because they see their operations as engineering-style manufacturing rather than 'food processing'. The crux of the matter is that you need real-time information, says Everington. "You need to be connected to the machines to record performance accurately. You need to involve the operators in determining what the problems are so that you get accurate reasons for downtime. You need to be able to track not just the performance of the machines, but what product they are making and therefore what rate they should be running at. You need to feed back in the scrap figures. And all this is what we do as part of the SCADA [supervisory control and data acquisition] software we have developed."

Over the years PLCs have become ubiquitous and methods of connecting to them and extracting the data have become easier and easier. As a result many automation companies have fallen into the trap of moving more and more to PLCs to get the raw data. But, unwittingly, they were removing the operator from the loop. Moving to PLCs enabled a process to be fully automated, but the quality of the information you got back was not as good, argues Everington.

Operator involvement

A fault code from a PLC might detect that a carton has not been formed correctly, say, but it won't tell you what caused it, he says. "What you need is to put a small terminal on the machine to allow the operator to identify the problem from a set of possible reasons. For instance: 'It's the wrong stock in; the stock's upside down; the stock's damp; the corner's bent; it's a different thickness'. There is a mass of different reasons the operator can interpret that the machine can't possibly know."

One advantage of this approach, says Everington, is that you don't need to connect to the PLC at all. "All we need from the machine is a single pulse signal to identify that the machine is running or stopped. When it's running we can count pulses and equate that to products being made. And when the pulses stop we can say that the machine has stopped and we start registering downtime." ​What you end up with is a system that allows the operator some free reign to put in a real reason for a stoppage. As a result you get a very accurate timing of the stoppage, you get an accurate rate at which your machines are running. And all of this is derived from one single pulsed signal.

"Because they have all got those pulsed signals, you're not relying on achieving communications with a PLC where the original vendor or manufacturer of the PLC is no longer around. The old problem of linking in to all the old bits of kit that are kicking around in many factories disappears."

Related news

Follow us

Featured Jobs

View more

Webinars

PRODUCTS & SERVICES