Milieu/Environment: The laboratory. It must be compliant in order to obtain repeatable and reproducible results. A good example of a non-compliant environment is a laboratory that is not climate controlled or is poorly climate controlled. Most analyses are sensitive to temperature variations. It makes little sense to invest in a laboratory to ultimately obtain poor results due to ineffective control of environmental conditions.
Material: The sample to be tested. The method of taking a representative sample would undoubtedly deserve a complete and additional note. The challenge is that it must be representative of the quality of the batch tested. The smaller the sample to be tested (a few grams), and the larger its source (a few tons), the more care must be taken in sampling. Once the sample has arrived at the laboratory, it must be maintained in the proper conditions (temp, humidity, etc.) if it is not tested immediately, and it must not be tested directly out of refrigeration, for example.
Machines: The apparatus that will be used to obtain the results. It will be necessary to make sure that this is well suited to carry out the measurement with the required precision (the choice of equipment is important). It is then necessary to ensure constant monitoring with external reference samples used to establish customized control charts that make it possible to check that the apparatus is correctly calibrated and that there is no deviation in the results. It is important to prepare a preventive maintenance schedule in order to avoid deviations in results which are sure to occur and to avoid service interruptions due to breakdowns from wear and tear.
Method: very often this is the standard (see the article "Standardization"). If there is no standard, it will be a user manual. Whatever the case may be, it is necessary to follow the method and not allow "adaptations", to not pass "tricks" from one operator to another, because the more the quality of the result depends on the know-how of the operator, the more the company takes risks in the event of his/her departure. If it is decided to use an "in-house" method (which is quite possible if it is properly described and, therefore, feasible in the same way by several operators), this method must make sense. For example, it makes no sense to carry out rheological analyses on flour obtained from dry, unconditioned wheat (see "Preparation of wheat for grinding").
Manpower: the person(s) who performs the analysis. The bottom line is that he/she must be properly trained by experts. This will have the advantage of preventing technicians from "sharing" bad habits or teaching others to ignore the method.
This may all seem very rigid, but we must remember the challenges of effective quality control. Even if it only involves recording figures in a notebook without further repercussions, that could still be conceived. But if the objective is to ensure that a flour will make it possible to obtain a finished product that makes customers stay loyal to a brand, then the financial stakes involved argue for rigor.
Finally, it must be emphasized that quality control is a voluntary and dynamic process. Too often we meet with technicians who say that they implement control “out of obligation.” To be clear, if quality control is seen as an obligation, it is because it is not fulfilling its role. And very often it is because it is not based on the finished product quality dynamic à raw material (described in the previous paragraph). There may have been a time when this feeling of performing analyses out of obligation was true. At a time when quality control was more often viewed as "avoiding trouble" instead of as a way to stand out. This is no longer true today, provided that you think differently and establish quality control based on the benefits it can bring to the business rather than constraints on suppliers.