Self-awareness facilitates a proper assessment of costconstrained cyber-physical systems, allocating limited resources where they are most needed. Together, situation awareness and attention are key enablers for self-awareness in efficient distributed sensing and computing networks.
The rapid development of intelligent control technology has also brought about changes in the automotive industry and led to development of autonomous or self-driving vehicles. To overcome traffic and environment issues, self-driving cars use a number of sensors for vision as well as a navigation system and actuators to control mechanical systems and computers to process the data. All these points make a self-driving car an interdisciplinary project that requires contribution from different fields. In our particular case, four different university departments and two companies are directly involved in the self-driving car project. The main aim of the paper is to discuss the challenges faced in the development of the first Estonian self-driving car. The project implementation time was 20 months and the project included four work packages: preliminary study, software development, body assembly and system tuning/testing of the self-driving car. This paper describes the development process stages and tasks that were distributed between the sub-teams. Moreover, the paper presents the technical and software solutions that were used to achieve the goal and presents a self-driving last mile bus called ISEAUTO. Special attention is paid to the discussion of safety challenges that a self-driving electrical car project can encounter. The main outcomes and future research possibilities are outlined
The demand for object detection capability in edge computing systems has surged. As such, the need for lightweight Convolutional Neural Network (CNN)-based object detection models has become a focal point. Current models are large in memory and deployment in edge devices is demanding. This shows that the models need to be optimized for the hardware without performance degradation. There exist several model compression methods; however, determining the most efficient method is of major concern. Our goal was to rank the performance of these methods using our application as a case study. We aimed to develop a real-time vehicle tracking system for cargo ships. To address this, we developed a weighted score-based ranking scheme that utilizes the model performance metrics. We demonstrated the effectiveness of this method by applying it on the baseline, compressed, and micro-CNN models trained on our dataset. The result showed that quantization is the most efficient compression method for the application, having the highest rank, with an average weighted score of 9.00, followed by binarization, having an average weighted score of 8.07. Our proposed method is extendable and can be used as a framework for the selection of suitable model compression methods for edge devices in different applications.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.