The proceedings contain 16 papers. The topics discussed include: artificial intelligence for the future of construction;cobots and industrial robots;predictive maintenance for wind turbine bearings: an MLOps approach ...
The proceedings contain 16 papers. The topics discussed include: artificial intelligence for the future of construction;cobots and industrial robots;predictive maintenance for wind turbine bearings: an MLOps approach with the DIAFS machine learning model;development of an artificial intelligence tool and sensing in informatization systems of mobile robots;PCA-NuSVR framework for predicting local and global indicators of tunneling-induced building damage;design and deployment of data development toolkit in cloud manufacturing environments;research and development of image processing algorithms for effective recognition of various gestures in real time;machine learning models for the recognition of commands in smart home technologies;responsive dehydration: sensor-driven optimisation of production cycles in a solar dehydrator;and formation of the method of description and control of the relative position of the links of the upper limbs of the grip of an anthropomorphic robot.
This paper reviews the application of deep learning in cell classification detection, highlights the importance of convolutional neural networks in image analysis, and discusses the application of image preprocessing,...
详细信息
Node deployment is a key issue in building sensor/actuator networks. To improve coverage and reduce costs, the use of autonomous mobile robotic nodes is an attractive option. This paper presents two new virtual force-...
详细信息
In this paper, we describe a bouncing strategy (smart strategy) for a mobile robot that uses one bit of memory for feedback, and guarantees that the robot will traverse all the rooms (and doorways) of a 2D environment...
详细信息
ISBN:
(数字)9781665468589
ISBN:
(纸本)9781665468596;9781665468589
In this paper, we describe a bouncing strategy (smart strategy) for a mobile robot that uses one bit of memory for feedback, and guarantees that the robot will traverse all the rooms (and doorways) of a 2D environment. The environment is modeled as a rectilinear polygon (also called orthogonal polygon), and the rooms and the doorways are defined by the decomposition algorithm we describe. Such a decomposition helps the robot to not go back to a room after leaving. We also define the notion of "virtual doors" that have the ability to let the robot through, or make the robot bounce from them. We compared three different types of bouncing rules: smart, random, billiard. The smart strategy grantees to reach to target. Although the random strategy on average behaves the same as the smart strategy, there are rectilinear polygons in which the robot cannot reach the target in the expected time steps. On the other hand, the billiard bouncing strategy can cause the robot to become trapped.
This paper introduces a novel design for flapping wing micro aerial vehicles (FWMAVs) that employs dynamic amplification. By utilising mechanical resonance, the design amplifies wing stroke and pitching motions, reduc...
详细信息
This paper studies continuous-Time model-free adaptive control (MFAC) framework for solving set-point tracking problems of second-order nonlinear time-invariant plants. First, the dynamical linearization process for c...
详细信息
Modeling multi-modal high-level intent is important for ensuring diversity in trajectory prediction. Existing approaches explore the discrete nature of human intent before predicting continuous trajectories, to improv...
详细信息
ISBN:
(纸本)9781728196817
Modeling multi-modal high-level intent is important for ensuring diversity in trajectory prediction. Existing approaches explore the discrete nature of human intent before predicting continuous trajectories, to improve accuracy and support explainability. However, these approaches often assume the intent to remain fixed over the prediction horizon, which is problematic in practice, especially over longer horizons. To overcome this limitation, we introduce HYPER, a general and expressive hybrid prediction framework that models evolving human intent. By modeling traffic agents as a hybrid discrete-continuous system, our approach is capable of predicting discrete intent changes over time. We learn the probabilistic hybrid model via a maximum likelihood estimation problem and leverage neural proposal distributions to sample adaptively from the exponentially growing discrete space. The overall approach affords a better trade-off between accuracy and coverage. We train and validate our model on the Argoverse dataset, and demonstrate its effectiveness through comprehensive ablation studies and comparisons with state-of-the-art models.
Since rolling bearing plays a crucial role in rotating machinery, the fault diagnosis research of it has been widely concerned in recent years. However, the collected fault signals often contain a lot of noise interfe...
详细信息
This paper investigates the application and benefits of using Generative Artificial Intelligence (GenAI) techniques, with an emphasis on large language models (LLMs) as an innovative approach in the development of inf...
详细信息
ISBN:
(纸本)9783031782657;9783031782664
This paper investigates the application and benefits of using Generative Artificial Intelligence (GenAI) techniques, with an emphasis on large language models (LLMs) as an innovative approach in the development of information systems. The focus lies primarily on systems employing open datasets where data confidentiality is not a focal concern. By replacing the traditional programmed logic with these advanced AI models, this paper presents the measurable improvements in the efficiency and adaptability of system design, development, and maintenance. The key advantages discussed include an enhanced rate and accuracy in query processing and response provision. Despite accentuating these substantial benefits, the paper acknowledges potential limitations with the alignment of these models to certain user-specific needs. A noteworthy component of this study includes an in-depth exploration of Large Language models usage and its implications in system interactivity and personalisation realms. The results suggest that the implementation of LLMs not only ensures precise and contextually relevant responses to user queries, but it also contributes to a significant reduction in time and resources expended during system design and development phases. Thus, replacing standard programmed logic with LLMs in systems utilizing open datasets accelerates the development timeline, simplifies maintenance procedures and boost the overall dynamism of the system.
Robots deployed in many real-world settings need to be able to acquire new skills and solve new tasks over time. Prior works on planning with skills often make assumptions on the structure of skills and tasks, such as...
详细信息
ISBN:
(纸本)9781728196817
Robots deployed in many real-world settings need to be able to acquire new skills and solve new tasks over time. Prior works on planning with skills often make assumptions on the structure of skills and tasks, such as subgoal skills, shared skill implementations, or task-specific plan skeletons, which limit adaptation to new skills and tasks. By contrast, we propose doing task planning by jointly searching in the space of parameterized skills using high-level skill effect models learned in simulation. We use an iterative training procedure to efficiently generate relevant data to train such models. Our approach allows flexible skill parameterizations and task specifications to facilitate lifelong learning in general-purpose domains. Experiments demonstrate the ability of our planner to integrate new skills in a lifelong manner, finding new task strategies with lower costs in both train and test tasks. We additionally show that our method can transfer to the real world without further fine-tuning.
暂无评论