The Industrial Internet of Things (IIoT) leverages thousands of interconnected sensors and computing devices to monitor and control large and complex industrial processes. Machine learning (ML) applications in IIoT use data acquired from multiple sensors to perform tasks such as predictive maintenance. While remembering useful learning from the past, these applications need to adapt learning for evolving sensor data stemming from changes in industrial processes and environmental conditions. This paper presents a continual learning pipeline to learn from the evolving data while replaying selected parts of the old data. The pipeline is configured to produce ML experiences (e.g., training a baseline neural network model), improve the baseline model with the new data while replaying part of the old data, and infer/predict using a specific model version given a stream of IIoT sensor data. We have evaluated our approach from an AI Engineering perspective using three industrial case studies, i.e., predicting tool wear, remaining useful lifetime, and anomalies from sensor data acquired from CNC machining and broaching operations. Our results show that configuring experiences for replay-driven continual learning allows dynamic maintenance of ML performance on evolving data while minimizing the excessive accumulation of legacy sensor data.

Abstract:Reviewing dictionaries is part of the ongoing work in lexicography, and several lexicographers have discussed the process and guidelines for reviews published in academic journals.However, few have addressed the evaluation of the outside matter and, if so, only in a cursory way. This article examines the evaluation of the outside matter in reviews published in Lexikos with a view to proposing some general principles for reviewing outer texts in printed and electronic dictionaries. The study shows that reviewers define the review object differently, some excluding the outside matter altogether, and that the way in which the outside matter is assessed differs within and between reviews. It is proposed that the separate sections of dictionaries should not only be examined independently but that their relationship to each other should also be evaluated so as to represent faithfully the lexicographic elements, i.e. wordlist, front, middle and back matter, their organisation and presentation, as well as three underlying elements: the function(s), data types and structures of the dictionary. Focus on all these elements may result in dictionary reviews that are academically sound because they treat the dictionary as a true research object. Keywords: BACK MATTER, DICTIONARY REVIEWS, EXTRA-LEXICOGRAPHIC SEC-TIONS, FRONT MATTER, LEXICOGRAPHIC INFORMATION COSTS, LEXICOGRAPHIC SECTIONS, MIDDLE MATTER, OUTSIDE MATTER, OVERRIDING OBJECTIVE, PREFACES, SCHOLARLY WRITINGS, SUBJECT-FIELD SECTIONS, SUBSTANCE OVER FORM, USER GUIDES, WORDLISTS Opsomming: Die evaluering van die buitewerk in woordeboekresensies.Die resensering van woordeboeke is deel van die deurlopende werk in die leksikografie, en verskeie leksikograwe het die metode en riglyne vir resensies wat in akademiese tydskrifte gepubliseer is, bespreek. Min het egter die evaluering van die buitewerk in resensies aangeroer, en, indien wel, slegs op 'n terloopse manier. Hierdie artikel ondersoek die evaluering van die buitewerk in resensies wat in Lexikos gepubliseer is, met die doel om 'n aantal algemene beginsels voor te stel vir die resensering van buitetekste in gedrukte en elektroniese woordeboeke. Die studie toon dat resensente die resensieobjek verskillend definieer, met sommige wat die buitewerk heeltemal uitsluit, en dat die manier waarop buitewerk beoordeel word, verskil binne en tussen resensies.Daar word voorgestel dat die aparte afdelings van woordeboeke nie alleen onafhanklik ondersoek behoort te word nie, maar dat hulle verhouding tot mekaar ook beoordeel behoort te word om die leksikografiese elemente getrou te verteenwoordig, d.w.s. woordelys, voor-, middel-en agterwerk, hul rangskikking en aanbieding, sowel as drie onderliggende elemente: die funksie(s), datatipes en strukture van die woordeboek. Fokus op al hierdie elemente kan lei tot woordeboekresensies wat akademies grondig is omdat hulle die woordeboek as 'n werklike navorsingsobjek behandel.