Preemptive Anomaly Prediction in IoT Components Alhassan Boner Diallo, Hiroyuki Nakagawa and Tatsuhiro Tsuchiya Osaka University, Osaka, Japan Abstract The Internet-of-Things (IoT) has become a very promising and fruitful area of research. The rapid development of IoT is revolutionizing our daily utilization of technology in every way. The IoT paradigm is that the devices making up an IoT system have resource constraints such as storage, computing and energy consumption. That paradigm makes possible a fexible and pervasive communication between devices that are bound to low resources. These constraints may create a state where there is anomaly occurrence on the component level that may impact the whole system. Some innovative techniques have been proposed to quantify the reliability of these devices for the aforementioned constraints. However, there is a gap between the quantifcation of the component reliability and the predictive and preemptive maintenance of these components. In this study, we propose an approach combining reliability quantifcation and reinforcement learning to build a mechanism that can achieve a predictive maintenance for the components of an IoT system such as devices and links. In the approach, a component-level mechanism is built to synthesize the reliability data, and to determine the probability of anomaly occurrence for each component. The approach is being applied to a self-adaptive IoT system for smart environment monitoring named DeltaIoT. Keywords self-adaptive systems, IoT, preliability, reinforcement learning, q-learning 1. Introduction Recently, the Internet of Things (IoT) has been one of the fastest growing felds in the computing domain. Its paradigm has been applied to many critical applications such as early warning systems for earthquake or tsunami, smart home security, trafc management, healthcare, and education systems, etc. Despite a rapid development and improvement in the IoT research area, many challenges remain. The challenges faced in IoT are related mainly to the following properties: scalability, availability, reli- ability, interoperability, security, mobility, performance, etc. The IoT infrastructure is made up of low resource devices, meaning that they have low storage and low computing power compared to other devices within the computing domain. This is the result of the desire to ac- commodate the energy consumption as most of the com- ponent rely on battery to power them up[1][2]. Nowa- days, the IoT paradigm is applied to many mission-critical systems, such as factory management, personal body sen- sors in healthcare, surveillance systems in nuclear power plants. These areas of application require a failure free system; otherwise there will be disastrous consequences. We must be able to trust these systems in all conditions as they impact the way we make numerous decisions CASA: 4th Context-aware, Autonomous and Smart Architectures International Workshop, ECSA’21 15-17 September 2021  a-bonerdiallo@ist.osaka-u.ac.jp (A. B. Diallo); nakagawa@ist.osaka-u.ac.jp (H. Nakagawa); t-tutiya@ist.osaka-u.ac.jp (T. Tsuchiya)  0000-0001-5280-4113 (H. Nakagawa) © 2021 Copyright for this paper by its authors. Use permitted under Creative Commons License Attribution 4.0 International (CC BY 4.0). CEUR Workshop Proceedings http://ceur-ws.org ISSN1613-0073 CEUR Workshop Proceedings (CEUR-WS.org) based on the data they collect and provide. The relia- bility of the IoT systems depends on the reliability of the components that make up the system. As the IoT devices are constrained by nature, there must be some mechanism in place to ensure their reliability at all time, in order to have accurate decision models based on the data provided by the lower layer of the IoT architecture. IoT reliability is a critical domain of research that has seen a lot of important contributions over the years. Mul- tiple ways of quantifying the reliability of IoT compo- nents have been proposed. However, there is a gap be- tween that quantifed reliability and its application in pre- dictive maintenance. In other words, how can we predict an accurate maintenance date for IoT components, based on the reliability measurement? To achieve that, we must build frst mechanisms that can synthesize the reliabil- ity information from anomalies to determine whether the system has become less reliable from that anomaly occurrence. The ability to reason about the quantifed reliability of the IoT system is a valuable step towards achieving predictive maintenance. The idea here is to build a dynamic decision-making process that can collect reliability data in a periodic manner and try to estimate a future failure time. Fundamentally, we can defne reliability as the study of failures. The reliability of a system or a computing device is its quality over a certain period of time. To quantify the reliability of a system or computing device, we use standard metrics all related to time like Mean Time To Failure, Mean Time Between Failures, and Mean Time To Repair, etc. Quantifying reliability is essential to assessing the continued success in the operation of an information system or a computing device.