Understanding and Monitoring your Manufacturing Process
There are two core issues which are essential for any business:-
- Produce a consistent, high quality product which meets the consumer’s needs
- Ensure that operations are functioning within their cost limits and that opportunities for improvement are constantly sought.
Every business is under a pressure of one kind or another – even if it doesn’t realise it. The most obvious sources of pressure are commercial (generally sales, profitability and cash flow) and capacity (ability to supply the market with the products it needs). There also are less common pressure sources, for example, availability of specific raw materials or quality issues.
Whilst other aspects of operations (for example marketing and distribution costs) will affect these pressures to a degree, a major factor is the performance of manufacturing process itself, which converts ideas and ingredients into finished products for distribution and sale.
If a business is to understand and respond these pressures, it must have an effective means of monitoring the manufacturing process and understanding the issues that affect efficiency, output and costs. This applies equally to small and larger businesses, but scale will affect the ways in which controls are designed and operated and the resources and skills available for monitoring.
Whilst raw data can be obtained in a number of ways depending on the parameters being measured – some of which will inevitably, in the confectionery industry, be subjective – converting that raw data into useful information which can be used to predict and correct aspects of the manufacturing operation requires a degree of understanding of sampling plans and statistical treatment of results. Too often data is not correctly interpreted and as a result opportunities are missed
Because of the cost of testing and measuring facilities and procedures there is inevitably a reluctance to undertake more than is necessary and often an inclination to construct conclusions “by eye” rather than dispassionately and statistically. The quality of the analytical support in the context of speed of response and consistent accuracy are paramount in building confidence in the function. It is also important for any analytical service to “sense check” results before publication and edit any obviously anomalous results.
Whilst many control parameters in the food industry are based on analytical testing (moisture content, fat content, etc.) there are few food products and processes which do not require sensory testing.
Analytical test results are reasonably easy to interpret – is a result within accepted limits – and charts can be used to illustrate trends, set levels at which corrective action is needed (“action limits”), etc. As long as the frequency of sampling, sampling method and handling, and analytical accuracy are defined and conformed to then the results will be useful and reliable.
Sensory testing can be much more difficult to obtain useful results and conclusions from, yet it is in many ways more important for product quality and consumer acceptability than analytical testing. This is because sensory testing relates directly to what the consumer wants and is expecting for the product and it brings together all the multiple facets of the product’s delivery.
A useful and effective sensory profiling “system” can range from something simple such as relatively unstructured tasting (preferably in a group) to much more sophisticated profiling using specially trained tasters in very controlled conditions. The cost and complexity of the latter approach is such that it is really only available to large organisations. It also has the drawback of (generally) a relatively long response time.
It is perhaps important to think clearly what we are trying to achieve:-
- Ensure the final product delivers to the consumer consistent satisfaction and that unsatisfactory product is identified quickly and reliably.
- Any organoleptic issues at critical intermediate steps are quickly identified so that corrective action can be taken and unsatisfactory product which has to be scrapped or recycled is minimised
- Raw and packaging materials are of the required quality and do not have “off” flavours or excessive odour levels
- Product is acceptable in quality throughout its defined shelf life
The real answer here is to have a group of reliable, responsive individuals who understand and have experience of the product. It is also important that one person’s views do not become pre-eminent – all people are equal when it comes to tasting.
Item 1 needs company wide involvement and in many cases testing can be undertaken at two levels. In small businesses it may be necessary to rely on a simple approach based on individual’s perceptions, but even then an occasional considered discussion is invaluable.
- Daily, individual tasting by as large a group as possible of whatever products are in production. This should certainly include staff from every area of the business – if accountants taste products more often they will understand much more about issues like changes in raw materials and the effects of potential cost saving measures!
- Regular group tasting in panels, preferably with a degree of training of members and use of “standard” samples for comparison. This should again include people from all areas and it should be regarded as a genuinely important function.
Item 2 in particular needs “immediacy” in order to rapidly identify and correct issues which may lead to rejection and scrapping of finished product. Although eating product in the production environment is rightly frowned upon, the most rapid and often the most aware QC comes from the operators and supervisors in the environment. If a system can be arranged whereby these people have a strong but ordered role in QC it is not only advantageous for the product, it enhances commitment and accountability in the manufacturing team.
Item 3 is more of a “back office” QC function but one which should not be neglected. Unsatisfactory raw or packaging materials must not be allowed to enter the production environment. Incoming ingredients and packaging materials must be sampled, examined and approved positively before use. It is also useful to retain samples of raw materials for a period in case some issue comes to light at a later date.
Item 4 is often forgotten but is actually very important. The consumer rarely if ever sees genuinely fresh product, it has generally had to pass through the supply chain and may be some weeks old. For this reason retained samples – which will probably have been kept under ideal conditions – should be tasted regularly. In addition samples should be uplifted from trade and consumer outlets on a reasonably regular basis so that quality as seen by the consumer can be audited.
Sampling and Understanding raw and packaging material quality
Efficient manufacture of quality products demands control and monitoring of raw and packaging materials to ensure predictable operations and product quality. Natural materials will always have a degree of variation which has to be accepted and understood. It is important to understand what degree of variation is acceptable and to develop systems which allow an understanding of raw material properties and how they will affect manufacturing efficiency and product quality.
A properly designed sampling plan and statistical treatment of results where appropriate will convert even subjective raw data, into results and information which will allow a proper interpretation of variations in quality and the potential effects on your process and product. It will also help you to establish trends which may allow pre-emptive action or response to suppliers, avoiding “drift” and possible scrap and rejection in the future.
In addition to an effective sampling plan it is important to have effective sampling procedures and systems which ensure samples are analysed and ingredients and part processed materials clearly marked as cleared (or rejected) before they are released for production.
Because of the specialist nature of some materials (for example sophisticated fats) businesses frequently accept the analytical data from suppliers’ laboratories which are capable of specific analytical tasks. Whilst this is a sensible and economic approach, particularly for small companies, there should be a check procedure in place to ensure that the results supplied are accurate. It may also be necessary to specify tests via a third party which a supplier may not normally undertake but which could have an impact on a particular process or product.
Understanding raw material yield and losses
Maximising the product yield from the raw materials that a business procures is the first step in cost control and manufacturing efficiency. In order to do this a systematic approach is needed relating output, recycle and losses to inputs
Properly designed and constructed mass balances allow a set of standard figures to be constructed and allowance for losses to be introduced as a variable at each stage. The balance needs to be constructed with proper mathematical rigor so that it can highlight operational faults and identify points for investigation.
Mass balance construction, manipulation and use is a sublect which needs its own treatment and will be covered in a subsequent article. It is not a complex procedure, rather a systematic approach which in itself can improve understanding of the process.
Measuring and monitoring process conditions
Effective control and monitoring of process conditions at every step of the manufacturing process is central to efficient operations and quality production. Inevitably there will be a degree of routine variation but the degree of variation which is acceptable at each stage has to be established and trends monitored so that corrective action can be taken when appropriate
Modern process control systems are extremely powerful, but not all businesses can afford and properly support them and a simpler system based on more direct involvement of floor staff using log sheets and control charts can be much more appropriate and cost effective.
There are a number of issues to overcome in achieving proper process monitoring:
- Accurate measurement – obtaining accurate and consistent data can be difficult in the food industry due to the nature of the materials being processed. Issues such as probe fouling can change the accuracy of a measurement over time in an unpredictable manner so it is important to have confidence in the measurement technology and installation being used.
- Proper recording – most modern process control systems record data regularly and in a way which is easy to access and process. If manual recording is used then a clear protocol needs to be established for each measurement to try to ensure consistency and a properly constructed log sheet used. Some manual recording can be a very useful way of ensuring that operators and managers are properly involved in the process are making appropriate inspections and checks.
- Interpretation – results need to be interpreted correctly and trends established using charts and data processing so that corrective action can be identified and taken in an appropriate manner. Excessive intervention can actually make process control less effective and lead to “hunting” in the system.
Understanding and monitoring yields and losses in process
There are inevitable “in process” losses – ranging from simple changes in moisture content to carry over and losses from washing and product changes. It is important to understand and monitor each of these as they relate to various aspects of process operation and ingredient characteristics so that there is a full understanding of yield losses.
Mass balances as mentioned above allow predicted values for losses to be calculated and comparisons made with actual values. Overall yield – converting as much as is possible of purchased ingredients into finished product is the ultimate measurement but it is important to understand where losses are occurring.
Possible areas for yield loss can include:-
- Excess moisture or byproduct in raw materials
- Spillage and other losses in storage and “in plant” transfers
- Inaccurate weighing and measuring of ingredients into process
- Deviation from recipe
- Losses due to:-
- Plant and process cleaning
- Out of quality product not recycled
- Product change over
- In process scrap
- Product out of specification
- Poor weight control in finished product
- Failure to wrap every good product
- Unrecorded product losses during shift change
There are many more potential loss areas specific to individual plants and installations, but this list shows the need to properly monitor the process operations and weight control of ingredient usage and finished product.
Monitoring in process and finished product quality
Consistent finished product quality is a core requirement in consumer satisfaction and comes directly from controlling ingredient and “in process” quality. It is essential to have systems which monitor quality in an objective manner and which provide useful information about trends in quality at every stage of the process.
There will be three quite separate areas of product and in process quality monitoring and the need to be approached and analysed differently:-
- Analytical data such as moisture or fat content, particle size, pH, etc. These are objective measurements and here the key is accuracy, consistency and speed coupled with proper interpretation and planned response to out of standard results
- Organoleptic data based on formal or casual tasting of raw materials, intermediates and finished product. These are subjective measurements which need a carefully constructed set of descriptors relevant to the product and its characteristics. There is also the need for a scoring system and reference back to a standard. Although in many ways the most important set of data (because it reflects potential consumer response) it is also the most difficult to set up and use, particularly in a small company where numbers of potential tasters are limited.
- Physical appearance of the carton, wrapped and unwrapped products. Although somewhat subjective these issues are relatively easy to define and inspect for.
Understanding Sensory Evaluation
Whilst a full sensory evaluation protocol may be beyond the resources of small and even medium sized businesses, a good understanding of the system is useful. Even if only used occasionally, it can be invaluable in increasing confidence in the product and manufacturing process as well as constantly improving the product to get close to the consumers’ ideal. It is also very useful for examining ex market product to understand the keeping properties of the product under “real life” conditions.
The sensory properties of a confectionery product are arguably among the most important of its quality features, and therefore it is very important to monitor the sensory profile of a product to ensure supply will meet consumers’ expectations. A starting point for monitoring the sensory quality of a confectionery product is to develop a robust sensory specification.
The sensory specification of a product is a qualitative and quantitative description of the target sensory profile of the finished product. The specification may give an indication of which attributes are critical and how much the product may vary and still be considered within specification. The sensory specification is usually developed by merging information about the product concept, the actual sensory profile of standard production, analysis of common sensory issues with the product, normal batch variation, and consumer expectations of the product and acceptance of variations to it.
The sensory specification can be at the core of various monitoring systems such as quality control, quality assurance, process monitoring, etc. Simple sensory quality systems involve trained assessors evaluating batches of products and grading them against the sensory specification and/or a reference product. Good sensory quality systems will also include diagnostics that identify specific areas of concern when a product is not found to be within specification.
Assessors involved in sensory quality monitoring systems should be screened for relevant sensory acuities and trained in both the basic concepts of objective sensory evaluation and the specific protocols used in the company. These assessors should be made familiar with a wide range of products/batches and examples of in and out of specification material. It is important that evaluations remain objective, and sources of bias (such as information about samples, differences in sample presentation, etc.) are controlled as much as possible. In most cases it is advisable to use the consensus judgement from a panel of at least several trained assessors. Assessors’ performance should be routinely validated to ensure they are able to detect product that is out of specification and can identify the relevant quality issues in products. This may be done by routine checks with spiked or abused product.
Sensory quality measurements can be tracked over time. This may allow identification of trends or problems with processes, raw ingredients, etc. Tracking these measurements over time is also a way of establishing objective targets and measureable outcomes for product sensory quality.