intelligent functions need higher compute ability, and the deployment process is more complicated. In scenarios where power consumption, size, and weight are limited, neither CPU nor ASCI cannot meet computing power r...
详细信息
Carrying out research on the motion effect of electrical stimulation of pigeon robots can help us clarify the relationship between electrical stimulation parameters and motor responses in actual scenarios, which has i...
详细信息
Vehicular Cloud computing (VCC) offers a promising platform for supporting various automotive applications and services. However, its distributed and dynamic nature makes it susceptible to Denial-of-Service (DoS) atta...
详细信息
Artificial intelligence is currently the most popular technology. With the advent of the era of big data, information technology based on artificial intelligence has become a new type of technology capable of most of ...
详细信息
Advanced manufacturing systems increasingly rely on intelligent algorithms to discriminate, model and predict system behaviours that lead to increased productivity. Edge intelligence allows the industrial systems to c...
详细信息
ISBN:
(数字)9781665467469
ISBN:
(纸本)9781665467469
Advanced manufacturing systems increasingly rely on intelligent algorithms to discriminate, model and predict system behaviours that lead to increased productivity. Edge intelligence allows the industrial systems to collect, compute and act based on process data while reducing the latency and cost associated to an hierarchical control system in which complex decisions are generated in the upper layers of the automation hierarchy. Greater local computing capabilities allow the online operation of such algorithms while accounting for increased performance requirements and lower sampling periods of the control loops. In this work we present the concept of a cognitive robotic cell that collects, stores and processes data in situ for enabling the control of a robotic arm in a production setting. The main features that characterise the robotic cell are embedded computing, open interfaces, and standards-based industrial communication with hardware peripherals and digital twin models for validation. An application of part classification is presented that uses the YOLOv4 image processing algorithm for real-time and online assessment that guides the control of an ABB IRB120C robotic arm. Results illustrate the feasibility and robustness of the approach in a real application. Quantitative evaluation underlines the performance of the implemented system.
City travelers rely heavily on online cab booking services for their daily transportation needs, making them an essential component of everyday commutes. Taxi businesses are seeing an increase in demand for integrated...
详细信息
There is an enormous demand for automated assessment systems, which can handle the bulk assessments efficiently and effectively. While existing systems perform well in grading objective questions, they break down with...
详细信息
In wireless networks, interference is always present and traditional anti-jamming methods are no longer able to cope with the more intelligent and dynamic interference. Therefore, there is an urgent need to improve an...
详细信息
High voltage circuit breakers play an important dual role in equipment control and protection during power grid operation, with core components including arc extinguishing chambers, dynamic and static contacts, and op...
详细信息
Convolutional Neural Networks (CNNs) are extensively used in cyberphysical systems for tasks like object detection and semantic segmentation. Given the critical nature of these systems, the ability of CNNs to meet str...
详细信息
ISBN:
(纸本)9789819755905;9789819755912
Convolutional Neural Networks (CNNs) are extensively used in cyberphysical systems for tasks like object detection and semantic segmentation. Given the critical nature of these systems, the ability of CNNs to meet stringent timing constraints is as crucial as their accuracy. However, variability in CNN execution times can lead to time constraint violations, affecting system reliability. To address this, we introduce RetNAS, an efficient neural architecture search framework that combines a rapidWorst-Case Execution Time (WCET) estimator and a constraint schedule to ensure timely performance. We utilize extreme value theory to estimate WCET, leveraging a generalized Pareto distribution based on limited execution time samples. Furthermore, RetNAS employs a constraint schedule to accelerate the search efficiency, which pursues a gradual search trajectory. RetNAS significantly enhances search efficiency and achieves a success rate of approximately 99.2% in meeting time constraints, outperforming other methods. Our experiments also show that CNNs designed by RetNAS surpass the accuracy of manually designed CNNs and visual transformers by 0.4% to 5%.
暂无评论