Many industries today are struggling with early the identification of quality issues, given the shortening of product design cycles and the desire to decrease production costs, coupled with the customer requirement for high uptime. The vehicle industry is no exception, as breakdowns often lead to on-road stops and delays in delivery missions. In this paper we consider quality issues to be an unexpected increase in failure rates of a particular component; those are particularly problematic for the original equipment manufacturers (OEMs) since they lead to unplanned costs and can significantly affect brand value. We propose a new approach towards the early detection of quality issues using machine learning (ML) to forecast the failures of a given component across the large population of units. In this study, we combine the usage information of vehicles with the records of their failures. The former is continuously collected, as the usage statistics are transmitted over telematics connections. The latter is based on invoice and warranty information collected in the workshops. We compare two different ML approaches: the first is an auto-regression model of the failure ratios for vehicles based on past information, while the second is the aggregation of individual vehicle failure predictions based on their individual usage. We present experimental evaluations on the real data captured from heavy-duty trucks demonstrating how these two formulations have complementary strengths and weaknesses; in particular, they can outperform each other given different volumes of the data. The classification approach surpasses the regressor model whenever enough data is available, i.e., once the vehicles are in-service for a longer time. On the other hand, the regression shows better predictive performance with a smaller amount of data, i.e., for vehicles that have been deployed recently.

Predictive Maintenance (PM) is a proactive maintenance strategy that tries to minimize a system’s downtime by predicting failures before they happen. It uses data from sensors to measure the component’s state of health and make forecasts about its future degradation. However, existing PM methods typically focus on individual measurements. While it is natural to assume that a history of measurements carries more information than a single one. This paper aims at incorporating such information into PM models. In practice, especially in the automotive domain, diagnostic models have low performance, due to a large amount of noise in the data and limited sensing capability. To address this issue, this paper proposes to use a specific type of ensemble learning known as Stacked Ensemble. The idea is to aggregate predictions of multiple models—consisting of Long Short-Term Memory (LSTM) and Convolutional-LSTM—via a meta model, in order to boost performance. Stacked Ensemble model performs well when its base models are as diverse as possible. To this end, each such model is trained using a specific combination of the following three aspects: feature subsets, past dependency horizon, and model architectures. Experimental results demonstrate benefits of the proposed approach on a case study of heavy-duty truck turbochargers.

Ensemble learning methods combine multiple models to improve performance by exploiting their diversity. The success of these approaches relies heavily on the dissimilarity of the base models forming the ensemble. This diversity can be achieved in many ways, with well-known examples including bagging and boosting. It is the diversity of the models within an ensemble that allows the ensemble to correct the errors made by its members, and consequently leads to higher classification or regression performance. A mistake made by a base model can only be rectified if other members behave differently on that particular instance, and provide the aggregator with enough information to make an informed decision. On the contrary, lack of diversity not only lowers model performance, but also wastes computational resources. Nevertheless, in the current state of the art ensemble approaches, there is no guarantee on the level of diversity achieved, and no mechanism ensuring that each member will learn a different decision boundary from the others. In this paper, we propose a parallel orthogonal deep learning architecture in which diversity is enforced by design, through imposing an orthogonality constraint. Multiple deep neural networks are created, parallel to each other. At each parallel layer, the outputs of different base models are subject to Gram-Schmidt orthogonalization. We demonstrate that this approach leads to a high level of diversity from two perspectives. First, the models make different errors on different parts of feature space, and second, they exhibit different levels of uncertainty in their decisions. Experimental results confirm the benefits of the proposed method, compared to standard deep learning models and well-known ensemble methods, in terms of diversity and, as a result, classification performance.

We propose a two-stage surrogate-assisted evolutionary approach to address the computational issues arising from using Genetic Algorithm (GA) for feature selection in a wrapper setting for large datasets. The proposed approach involves constructing a lightweight qualitative meta-model by sub-sampling data instances and then using this meta-model to carry out the feature selection task. We define "Approximation Usefulness" to capture the necessary conditions that allow the meta-model to lead the evolutionary computations to the correct maximum of the fitness function. Based on our procedure we create CHC_QX a Qualitative approXimations variant of the GA-based algorithm CHC (Cross generational elitist selection, Heterogeneous recombination and Cataclysmic mutation). We show that CHC_QX converges faster to feature subset solutions of significantly higher accuracy, particularly for large datasets with over 100K instances. We also demonstrate the applicability of our approach to Swarm Intelligence (SI), with results of PSO_QX, a qualitative approximation adaptation of the Particle Swarm Optimization (PSO) method. A GitHub repository with the complete implementation is available². This paper for the Hot-off-the-Press track at GECCO 2023 summarizes the original work published at [3].