AES_8

CONTROLLO tecnica 96 Novembre - Dicembre 2021 n Automazione e Strumentazione Figure 3 - Time distribution histograms like the ones shown are automatically generated once the set-up is implemented. this type of output can be updated only if multiple shifts have occurred, thus preventing privacy violation Once the set-up is completed the devices start acquiring and saving data in their internal memory or sending it to a predetermined cloud server via wi-fi connection. Since no Machine Learning is involved, the system solely samples those statuses and operations that are described within the status list, i.e. the procedure or work-cycle. Exceptions are not considered. This way, the analysis can be used to determine what are the standard time and the level of control over each task, expressed as its vari- ability ( Figure 3 ). In the presented case, such measures have been used to determine how much time got wasted in non-value-adding activities and for comparison rea- sons between multiple benches’ designs and procedures. This, however, should not divert the attention from the necessary communication process. The introduction of new and innovative tools within manufacturing organi- zations should always be intended as an opportunity to enhance the dialogue between white- and blue-collars. If not accurately introduced to all the stakeholder, auto- matic systems like the one here presented lose most of their potential and risk to carry new tensions. Towards a multi-purpose tool The described approach [2] is currently being tested in multiple manufacturing companies of different sizes and with different objectives. Encouraging results have been achieved in cyclical assembly processes , where the need for efficiency and ergonomics measures is strongly felt. Positive feedbacks have also been col- lected from workers, whose attitude towards the auto- matic system and the wearable devices specifically has proven to be highly dependent by the communication process: as expected, where involved from the very first steps of the projects, people have kept the most proac- tive behavior. The need for both ergonomic measures and higher finesse of the detected activities has led to new devel- opments of the system, that is going to be tested as an automatic support tool for OCRA and MTM-UAS ana- lyses. n References [1] Manuale corso di applicatore MTM-UAS, Associ- azione MTM Italia. [2] The presented, patent pending, solution has been developed by Università di Pisa, Erre Quadro s.r.l and TOI s.r.l.. La rilevazione dei Tempi e Metodi nell’era dell’Internet delle cose Il miglioramento continuo presuppone un’in- tensa attività di campionamento delle prestazioni. Nell’ambito delle attività manuali le tecniche tra- dizionalmente utilizzate per la raccolta dati presen- tano limitazioni in termini di scalabilità, invasività ed oggettività del dato. La soluzione presentata mira a superare questi vincoli con l’introduzione di un sistema IoT automatico e flessibile in grado di raccogliere e analizzare dati non ambigui dai processi operativi. What about privacy? To assure the privacy of workers and prevent improper uses of data, a procedure has been developed that thanks to a randomization process that totally anonymizes the data acquired from multiple shifts making it impossi- ble to measure one single worker’s performance. This element perfectly matches with the objectives of the presented project, that focuses on processes’ and not people’s performance. By a legal point of view, in matter of privacy the art. 4 of Law 300/1970 (Workers’ Statute) has to be con- sidered, which represents the reference within Italian legislation. Under specific hypotheses – e.g. where on field data is already used to manage the production – automatic systems like the one here presented can be deemed as work tools necessary to perform the job.