Internet of Things (IoT) applications in several domains such as surveillance of civil infrastructure, smart grids, and smart healthcare are of utmost importance for our society and require dependable performance. Guaranteeing that application-specific dependability requirements are met is however still an open research challenge. The IoT indeed exposes highly resource-constrained computing devices to harsh environmental conditions (e.g., heat, mechanical shock, electromagnetic radiation) and physical attacks. Unfortunately, traditional methods to withstand these threats heavily rely on redundancy, a concept that is incompatible with the resource constraints of common IoT devices. In this article, we illustrate our efforts in providing methods and tools to predict, guarantee, and raise the level of dependability of the IoT. We first outline our contributions in the area of dependable wireless networking and describe a cost-effective solution allowing to guarantee that IoT applications meet specific performance requirements despite the challenging interaction of low-power wireless networks with their surrounding environment. We then argue that dependable networking alone is insufficient to guarantee the correct operation of a complex IoT system, and outline how we join different scientific disciplines in a long-term endeavor and work towards a coherent view of dependability.

Several ultrasound based localization systems consist of environmental anchor nodes, and mobile nodes, which estimate their own position by using a static infrastructure. For the location process, every anchor has to know its position. In most approaches, the location of all anchors has to be determined a priori manually. This procedure is time consuming and fault-prone. In this paper, we present Distribute & Erase and Explorer, two self-calibration methods for ultrasound based localization systems. The first uses a set of three pre-calibrated anchors to explore the whole localization system whereas the second refines the anchors' positions progressively. Both of our approaches require no additional hardware besides ultrasound receivers or transmitters, and radio transceivers, which have to be already available for the WSN based localization system.

Due to increasing complexity of software in embedded systems, performance aspects become much more important this days. This should happen early in the development process. Often execution times and events are not easily countable or measurable due to a lack of functionality in these systems. Execution time monitoring is also relevant in terms of reacting to internal and external events dynamically. Especially for systems using multiple tasks with internal or external resource dependencies, this is a major discipline. Another problem is that measurements during the development process are often done by interfering the system as a whole. This method leads to biases in the measurement results, because the finalized system gets deployed without these interfering functionalities and can therefore work more efficiently than the development system. The scope of the present work is to develop a module in a hardware description language (HDL) which is able to measure execution times and events task-aware and unaware without interfering the system. The measurements of this module must be handed to the programmer through an easy accessible interface. The main focuses of the project are the scalability, platform independency concerning processor and operating system (OS), as well as easy extendibility. Also, reusability of counters during runtime is included in this work.

The real-world implementation of a just theoretically elaborated idea is sometimes cumbersome, often a couple of obstacles have to be overcome. That's likewise in the area of wireless sensor networks (WSN), but complicated by some further restrictions, e.g. little memory or low power consumption. One well-known and often required application within WSN is the geographical localization of several sensor nodes. That's why this paper deals with some problems arising during the development of a WSN using the time difference of arrival (TDoA) of ultrasound and radio signals for positioning. Its focus is on handling of microcontroller difficulties like little memory, low computational power or low energy consumption as well as hardware driven failures like inaccurate measurements or node failures.

The use of formal methods to prove the correctness of compositional embedded systems is increasingly important. However, the required models and algorithms can induce an enormous complexity. Our approach divides the formal system model into layers and these in turn into modules with defined interfaces, so that reduced formal models can be created for the verification of concrete functional and non-functional requirements. In this work, we use Uppaal to (1) model an RTOS kernel in a modular way and formally specify its internal requirements, (2) model abstract tasks that trigger all kernel functionalities in all combinations or scenarios, and (3) verify the resulting system with regard to task synchronization, resource management, and timing. The result is a fully verified model of the operating system layer that can henceforth serve as a dependable foundation for verifying compositional applications w.r.t. various aspects, such as timing or liveness.

With the development of autonomous driving technology, the requirements for machine perception have increased significantly. In particular, camera-based lane detection plays an essential role in autonomous vehicle trajectory planning. However, lane detection is subject to high complexity, and it is sensitive to illumination variation, appearance, and age of lane marking. In addition, the sheer infinite number of test cases for highly automated vehicles requires an increasing portion of test and validation to be performed in simulation and X-in-the-loop testing. To model the complexity of camera-based lane detection, physical models are often used, which consider the optical properties of the imager as well as image processing itself. This complexity results in high efforts for the simulation in terms of modelling as well as computational costs. This paper presents a Phenomenological Lane Detection Model (PLDM) to simulate camera performance. The innovation of the approach is the modelling technique using Multi-Layer Perceptron (MLP), which is a class of Neural Network (NN). In order to prepare input data for our neural network model, massive driving tests have been performed on the M86 highway road in Hungary. The model’s inputs include vehicle dynamics signals (such as speed and acceleration, etc.). In addition, the difference between the reference output from the digital-twin map of the highway and camera lane detection results is considered as the target of the NN. The network consists of four hidden layers, and scaled conjugate gradient backpropagation is used for training the network. The results demonstrate that PLDM can sufficiently replicate camera detection performance in the simulation. The modelling approach improves the realism of camera sensor simulation as well as computational effort for X-in-the-loop applications and thereby supports safety validation of camera-based functionality in automated driving, which decreases the energy consumption of vehicles.