intelligent transportation systems (ITS) have gained significant traction since the 1980s and 1990s, driven by technological advancements and increasing urbanization, which have caused intense transportation challenge...
详细信息
This investigation introduces a novel deep reinforcement learning-based suite to control floating platforms in both simulated and real-world environments. Floating platforms serve as versatile test-beds to emulate mic...
详细信息
ISBN:
(纸本)9798350377712;9798350377705
This investigation introduces a novel deep reinforcement learning-based suite to control floating platforms in both simulated and real-world environments. Floating platforms serve as versatile test-beds to emulate microgravity environments on Earth, useful to test autonomous navigation systems for space applications. Our approach addresses the system and environmental uncertainties in controlling such platforms by training policies capable of precise maneuvers amid dynamic and unpredictable conditions. Leveraging Deep Reinforcement Learning (DRL) techniques, our suite achieves robustness, adaptability, and good transferability from simulation to reality. Our deep reinforcement learning framework provides advantages such as fast training times, large-scale testing capabilities, rich visualization options, and ROS bindings for integration with real-world robotic systems. Being open access, our suite serves as a comprehensive platform for practitioners who want to replicate similar research in their own simulated environments and labs.
Performing an inspection task while maintaining the privacy of the inspected site is a challenging balancing act. In this work, we are motivated by the future of nuclear arms control verification, which requires both ...
详细信息
ISBN:
(纸本)9798350377712;9798350377705
Performing an inspection task while maintaining the privacy of the inspected site is a challenging balancing act. In this work, we are motivated by the future of nuclear arms control verification, which requires both a high level of privacy and guaranteed correctness. For scenarios with limitations on sensors and stored information due to the potentially secret nature of observable features, we propose a robotic verification procedure that provides map-free exploration to perform a source verification task without requiring, nor revealing, any task-irrelevant, site-specific information. We provide theoretical guarantees on the privacy and correctness of our approach, validated by extensive simulated and hardware experiments.
Unmanned Aerial Vehicles (UAV) equipped with various sensors can efficiently and safely perform automated detection of overhead power lines. Deep learning-based image recognition technology plays a crucial role in thi...
详细信息
Driver fatigue is a leading cause of road accidents. This research presents a real-time driver drowsiness detection system using deep learning, optimized for low-power embedded systems. By compressing a complex model ...
详细信息
The emergence of the edge computing concept, such as in smart lighting controlsystems, is because the computing system proposed by the cloud concept often causes delays. The problem is that the end device of the Inte...
详细信息
ISBN:
(纸本)9783031477140;9783031477157
The emergence of the edge computing concept, such as in smart lighting controlsystems, is because the computing system proposed by the cloud concept often causes delays. The problem is that the end device of the Internet of things (IoT), such as NodeMCU as the device target for placing edge computing programs, has limited resources. This paper proposes a Quantized 8-bit K-Nearest Neighbor (Q8KNN), a novel quantization method that performs model compression on KNN with our case study, the smart lighting controlsystems, using NodeMCU. Firstly, we have created a novel and accurate smart lighting design that we proposed. The design uses edge computing and KNN models to predict the control data. Then we developed a quantization method to make the KNN model fit into the NodeMCU. Finally, we test the performance of our novel compression model using the Accuracy and Compression Ratio (CR) metrics. The test results show that the number of unique values that change due to quantization does not cause a cardinality problem. Then Q8KNN can provide CR up to 1.6 times. With a smaller model size, Q8KNN can increase the number of training samples in the NodeMCU memory. So, using the same model size, the original KNN gives an accuracy of 94% while Q8KNN can provide up to 98%.
This paper presents a method to recover quadrotor Unmanned Air Vehicles (UAVs) from a throw, when no control parameters are known before the throw. We leverage the availability of high-frequency rotor speed feedback a...
详细信息
ISBN:
(纸本)9798350377712;9798350377705
This paper presents a method to recover quadrotor Unmanned Air Vehicles (UAVs) from a throw, when no control parameters are known before the throw. We leverage the availability of high-frequency rotor speed feedback available in racing drone hardware and software to find control effectiveness values and fit a motor model using recursive least squares (RLS) estimation. Furthermore, we propose an excitation sequence that provides large actuation commands while guaranteeing to stay within gyroscope sensing limits. After 450ms of excitation, an Incremental Nonlinear Dynamic Inversion (INDI) attitude controller uses the 52 fitted parameters to arrest rotational motion and recover an upright attitude. Finally, a Nonlinear Dynamic Inversion (NDI) position controller drives the craft to a position setpoint. The proposed algorithm runs efficiently on microcontrollers found in common UAV flight controllers, and was shown to recover an agile quadrotor every time in live experiments with as low as 3.5m throw height, demonstrating robustness against initial rotations and noise. We also demonstrate control of randomized quadrotors in simulated throws, where the parameter fitting Root-Mean-Square (RMS) error is typically within 10% of the true value.
The research worked particularly on the situation where an early warning system fails to forecast the occurrence of an earthquake when the earthquake strikes at night, in a darkness, and in blackout state. As the resu...
详细信息
ISBN:
(纸本)9798350349603;9798350349597
The research worked particularly on the situation where an early warning system fails to forecast the occurrence of an earthquake when the earthquake strikes at night, in a darkness, and in blackout state. As the result, people are assumed to move randomly. Some would sit/stand still, the others would move slowly forward or to the left or to the right. Such a movement would scarcely reach the evacuation area. The research has designed a device, named as the Citizen Movement control System (CMCS), that uses the coordinate of the evacuation area, measures the coordinate of the device itself, and raises sirens and lights in a particular manner based on the result of the analysis of both coordinates. In such a case, the sirens and lights would guide the citizens to the evacuation area, and would potentially help the provision of the life support from first responders. The simulations suggest that the CMCS would handle citizens properly. In ideal circumstances, inside a square km area, all citizens could be evacuated within 400 seconds. This value is better than the 450 seconds, recommended by the European Centre on Prevention.
Dexterous telemanipulation is crucial in advancing human-robot systems, especially in tasks requiring precise and safe manipulation. However, it faces significant challenges due to the physical differences between hum...
详细信息
ISBN:
(纸本)9798350377712;9798350377705
Dexterous telemanipulation is crucial in advancing human-robot systems, especially in tasks requiring precise and safe manipulation. However, it faces significant challenges due to the physical differences between human and robotic hands, the dynamic interaction with objects, and the indirect control and perception of the remote environment. Current approaches predominantly focus on mapping the human hand onto robotic counterparts to replicate motions, which exhibits a critical oversight: it often neglects the physical interaction with objects and relegates the interaction burden to the human to adapt and make laborious adjustments in response to the indirect and counter-intuitive observation of the remote environment. This work develops an End-Effects-Oriented Learning-based Dexterous Telemanipulation (EFOLD) framework to address telemanipulation tasks. EFOLD models telemanipulation as a Markov Game, introducing multiple end-effect features to interpret the human operator's commands during interaction with objects. These features are used by a Deep Reinforcement Learning policy to control the robot and reproduce such end effects. EFOLD was evaluated with real human subjects and two end-effect extraction methods for controlling a virtual Shadow Robot Hand in telemanipulation tasks. EFOLD achieved real-time control capability with low command following latency (delay<0.11s) and highly accurate tracking (MSE<0.084 rad).
In Today scenario , Android is one of the most frequently used mobile operatings systems, thus it is a priority for advanced threat actors and hackers. Malicious code is often found in Android applications to which se...
详细信息
暂无评论