An Empirical Evaluation of Prediction by Partial Matching in Assembly Assistance Systems

Abstract

Industrial assistive systems result from a multidisciplinary effort that integrates IoT (and Industrial IoT), Cognetics, and Artificial Intelligence. This paper evaluates the Prediction by Partial Matching algorithm as a component of an assembly assistance system that supports factory workers, by providing choices for the next manufacturing step. The evaluation of the proposed method was performed on datasets collected within an experiment involving trainees and experienced workers. The goal is to find out which method best suits the datasets in order to be integrated afterwards into our context-aware assistance system. The obtained results show that the Prediction by Partial Matching method presents a significant improvement with respect to the existing Markov predictors.

Assembly Assistance System with Decision Trees and Ensemble Learning

“…On the other hand, the stochastic methods [ 14 , 15 ] and even the LSTM [ 16 ] yield an inferior prediction rate compared to the prediction techniques based on decision trees explored in this paper, as the training data did not contain all the possible assembly sequences, thus making those algorithms unsuitable in scenarios with new data. In contrast, the methods based on decision trees have a prediction rate of 100% since for any input they are able to produce an outcome.…”

“…In other approaches, a Markov model tracks the occurrence frequencies of multiple possible next states within the pattern history table and, thus, it can provide multiple choices for the next assembly step [ 13 , 14 ]. The PPM algorithm was also evaluated [ 15 ] as an assembly step predictor. It combines different order Markov predictors.…”

“…In [ 16 ], Precup et al predicted the next assembly step with an optimally configured LSTM neural network. Among all of the above mentioned assembly modelling techniques, the PPM [ 15 ] proved to be the most accurate.…”

“…The user has three interaction buttons: play / repeat instruction, previous or next instruction. More details about the current assembly assistance system’s implementation and its concept can be found in [ 15 , 49 ], respectively.…”

“…In the previous work [ 4 ], two-level context-based predictors have been studied for such a predictive assembly modeling. In [ 13 , 14 ], Markov predictors were investigated, whereas in [ 15 ] the Prediction by Partial Matching (PPM) algorithm was analyzed. In [ 16 ], a Long Short-Term Memory (LSTM) was involved to predict the next assembly steps.…”