Andreas Theissler

Hochschule Aalen

The analysis of time series data is of high relevance in fields like manufacturing, health, automotive, or science. In this paper, we propose ROCKAD, a kernel-based approach for semi-supervised whole time series anomaly detection, i.e. the assignment of a single anomaly score to an entire time series. Our key idea is to use ROCKET as an unsupervise...

In minimally invasive surgery (MIS), the reliable detection of hard inclusions in soft tissue is crucial for the success of the intervention. In robot-assisted surgery (RAS) however, limited technologies are available for intracorporeal tissue stiffness assessment due to the lack of force and tactile feedback from the robot tool tip. This paper inv...

In clinical practice, algorithmic predictions may seriously jeopardise patients’ health and thus are required to be validated by medical experts before a final clinical decision is met. Towards that aim, there is need to incorporate explainable artificial intelligence techniques into medical research. In the specific field of epileptic seizure dete...

Labeling of datasets is an essential task for supervised and semi-supervised machine learning. Model-based active learning and user-based interactive labeling are two complementary strategies for this task. We propose VisGIL which, using visual cues, guides the user in the selection of instances to label based on utility measures deduced from an ac...

Time series data is increasingly used in a wide range of fields, and it is often relied on in crucial applications and high-stakes decision-making. For instance, sensors generate time series data to recognize different types of anomalies through automatic decision-making systems. Typically, these systems are realized with machine learning models th...

For the manufacturing of miniature force/torque sensors, extreme accuracy is required due to the tiny size of the strain gauges inside the sensors (2×2.5 mm). The current method of manually assembling them by hand is difficult, timeintensive, and error-prone. To improve this, a system to pick up the tiny objects from a plate and place them on eleme...

In machine learning, the presumably best model is selected from a variety of model candidates generated by testing different model types, hyperparameters, or feature subsets. The advent of deep learning has made model selection even more challenging due to the huge parameter search space. Relying on a single metric to select the best model does not...

Current prototype-based classification often leads to prototypes with overlapping semantics where several prototypes are similar to the same image parts. Also, single prototypes tend to activate highly on a mixture of semantically different image parts. This impedes interpretability since the nature of the connections between the parts is unknown....

Recent developments in maintenance modelling fueled by data-based approaches such as machine learning (ML), have enabled a broad range of applications. In the automotive industry, ensuring the functional safety over the product life cycle while limiting maintenance costs has become a major challenge. One crucial approach to achieve this, is predict...

In the field of predictive maintenance (PdM), machine learning (ML) has gained importance over the last years. Accompanying this development, an increasing number of papers use non-interpretable ML to address PdM problems. While ML has achieved unprecedented performance in recent years, the lack of model explainability or interpretability may manif...

The estimation of a system's or a component's remaining useful life (RUL) is considered the most complex task in predictive maintenance, at the same time the most beneficial one. In this brief review paper, we survey the state-of-the-art in machine learning-based RUL prognosis based on research on NASA's C-MAPSS data set. We identify the frequently...

The human sense of touch allows recognizing a wide set of properties of a grasped object such as weight, shape, hardness, temperature or surface texture. Despite the great importance of haptic sensing for humans, mechatronic end-effectors of humanoid robots and industrial manipulators are rarely endowed with tactile feedback. This is due to a lack...

Human-centered machine learning is becoming an emerging field aiming to enable domain experts that do not necessarily have a data science background to make use of machine learning applications. Especially in unsupervised machine learning, e.g. cluster analysis, models cannot be autonomously tuned towards an optimal solution for a given application...

The labeling of datasets is an important task for supervised and semi-supervised machine learning that can be addressed with visual analytics. With model-based active learning and user-based interactive labeling, there are two complementary strategies for this task. We present an approach that combines the strengths of both areas and aims to guid...

One of the major problems of applying supervised machine learning methods in real-world problems is the absence of labeled data. Labeling huge amounts of data is time consuming and cost intensive. Moreover, in many cases, labels can only be assigned by domain experts like medical doctors or engineers, who have little time and do not necessarily hav...

Obtaining new information and creating value from present measurements without introducing additional sensors is cost-efficient and mitigates data that is collected and stored by information systems but not used. In electromechanical drive systems, defect states of synchronous motors can be detected based on measurements of the motor current. While...

In this paper we describe our approach to solve the text classification problem of the Chat Analytics for Twitch (ChAT) discovery challenge of ECML-PKDD 2020. The task was to predict the subscription status of Twitch users for a given channel based on their comments posted within the Twitch chat. Users have the opportunity to support channels in th...

In anomaly detection problems the available data is often not or not fully labelled. This leads to results that are usually significantly worse than in balanced classification problems. In this short paper VIAL-AD is proposed, which addresses this problem with a sequence of unsupervised, semi-supervised and supervised machine learning models allowi...

A major challenge during the development of Machine Learning systems is the large number of models resulting from testing different model types, parameters, or feature subsets. The common approach of selecting the best model using one overall metric does not necessarily find the most suitable model for a given application, since it ignores the diff...

Emotional intelligence plays an essential role in negotiations. In this work we present an approach for emotion recognition (sentiment detection) based on the characteristics of the human voice. It is shown how such an approach can support negotiators to train their emotional intelligence prior to negotiations. Using methods from the field of deep...

This paper introduces the modular anomaly detection toolbox OCADaMi that incorporates machine learning and visual analytics. The case often encountered in practice where no or only a non-representative number of anomalies exist beforehand is addressed, which is solved using one-class classification. Target users are developers, engineers, test engi...

Recent advancements in automotive technologies, most notably autonomous driving, demand electronic systems much more complex than realized in the past. The automotive industry has been forced to adopt advanced consumer electronics to satisfy the demand, and thus it becomes more challenging to assess system reliability while adopting the new technol...

The complexity of vehicles has increased over the last years and will continue to do so. Hence, repairs in repair shops become more and more complex and thereby time-consuming, where time to repair is a competitive factor. During repairs and servicing of vehicles in the independent aftermarket the data read-out using diagnostic testers is transferr...

The massive growth of data produced in the automotive industry by acquiring data during production and test of vehicles requires effective and intelligent ways of analysing these recordings. In order to detect potential faults, data from the in-vehicle network interconnecting vehicle subsystems is recorded during road trials. The complexity and vol...

In the automotive industry test drives are being conducted during the development of new vehicle models or as a part of quality assurance of series-production vehicles. Modern vehicles have 40 to 80 electronic control units interconnected via the so-called in-vehicle net- work. The communication on this in-vehicle network is recorded during test dr...

In the automotive industry test drives are conducted during the development of new vehicle models or as a part of quality assurance for series vehicles. During the test drives, data is recorded for the use of fault analysis resulting in millions of data points. Since multiple vehicles are tested in parallel, the amount of data that is to be analyse...

The one-class support vector machine “support vector data description” (SVDD) is an ideal approach for anomaly or outlier detection. However, for the applicability of SVDD in real-world applications, the ease of use is crucial. The results of SVDD are massively determined by the choice of the regularisation Parameter C and the kernel parameter of t...

In modern vehicles 40 to 80 electronic control units are interconnected via the in-vehicle network. During test drives the network communication is recorded in order to locate faults, resulting in a multivariate time series with millions of data points for each test drive. Hence, manually analysing each recording in great detail is not feasible. Th...

The identification and optimization of problem sections in the road network, as well as the estimation and optimization of the time of travel is a challenge of great interest in the field of traffic engineering. The economic damage caused by congested traffic is tremendous. Complex and application-specific models reconstructing the current traffic...

Modern vehicles contain a highly complex network of hardware and software subsystems . To be able to locate faults or to evaluate the behaviour of subsystems, the communication on the vehicle’s networks is being recorded by measurement systems – so called data loggers. This is for example done during road trials or Hardware-in-the-loop (HiL)-tests....

For current automotive measurement and simulation systems the test of safety or comfort functions is a challenging field since it is a key issue for an effective, reliable, and efficient quality assurance in automotive industry. Examples for these Driver Assistance Functions are vehicles exchanging knowledge about road conditions or a system that w...

Der Entwicklungsprozess von eingebetteten Systemen in der Automobilindustrie folgt oftmals den Regeln des V-Modells. Die Zerlegung des Gesamtsystems startet meist auf der Ebene „Fahrzeug“. Danach wird das System in Komponenten zerlegt. Eine mögliche Zerlegungskette ist die Aufteilung in Steuergeräte, Steuergeräte-Hardware/-Software und anschließend...