-
Agentic AI for Intent-Based Industrial Automation
Authors:
Marcos Lima Romero,
Ricardo Suyama
Abstract:
The recent development of Agentic AI systems, empowered by autonomous large language models (LLMs) agents with planning and tool-usage capabilities, enables new possibilities for the evolution of industrial automation and reduces the complexity introduced by Industry 4.0. This work proposes a conceptual framework that integrates Agentic AI with the intent-based paradigm, originally developed in ne…
▽ More
The recent development of Agentic AI systems, empowered by autonomous large language models (LLMs) agents with planning and tool-usage capabilities, enables new possibilities for the evolution of industrial automation and reduces the complexity introduced by Industry 4.0. This work proposes a conceptual framework that integrates Agentic AI with the intent-based paradigm, originally developed in network research, to simplify human-machine interaction (HMI) and better align automation systems with the human-centric, sustainable, and resilient principles of Industry 5.0. Based on the intent-based processing, the framework allows human operators to express high-level business or operational goals in natural language, which are decomposed into actionable components. These intents are broken into expectations, conditions, targets, context, and information that guide sub-agents equipped with specialized tools to execute domain-specific tasks. A proof of concept was implemented using the CMAPSS dataset and Google Agent Developer Kit (ADK), demonstrating the feasibility of intent decomposition, agent orchestration, and autonomous decision-making in predictive maintenance scenarios. The results confirm the potential of this approach to reduce technical barriers and enable scalable, intent-driven automation, despite data quality and explainability concerns.
△ Less
Submitted 5 June, 2025;
originally announced June 2025.
-
Regularização, aprendizagem profunda e interdisciplinaridade em problemas inversos mal-postos
Authors:
Roberto Gutierrez Beraldo,
Ricardo Suyama
Abstract:
In this book, written in Portuguese, we discuss what ill-posed problems are and how the regularization method is used to solve them. In the form of questions and answers, we reflect on the origins and future of regularization, relating the similarities and differences of its meaning in different areas, including inverse problems, statistics, machine learning, and deep learning.
In this book, written in Portuguese, we discuss what ill-posed problems are and how the regularization method is used to solve them. In the form of questions and answers, we reflect on the origins and future of regularization, relating the similarities and differences of its meaning in different areas, including inverse problems, statistics, machine learning, and deep learning.
△ Less
Submitted 14 February, 2025;
originally announced February 2025.
-
Intelligent Fault Diagnosis of Type and Severity in Low-Frequency, Low Bit-Depth Signals
Authors:
Tito Spadini,
Kenji Nose-Filho,
Ricardo Suyama
Abstract:
This study focuses on Intelligent Fault Diagnosis (IFD) in rotating machinery utilizing a single microphone and a data-driven methodology, effectively diagnosing 42 classes of fault types and severities. The research leverages sound data from the imbalanced MaFaulDa dataset, aiming to strike a balance between high performance and low resource consumption. The testing phase encompassed a variety of…
▽ More
This study focuses on Intelligent Fault Diagnosis (IFD) in rotating machinery utilizing a single microphone and a data-driven methodology, effectively diagnosing 42 classes of fault types and severities. The research leverages sound data from the imbalanced MaFaulDa dataset, aiming to strike a balance between high performance and low resource consumption. The testing phase encompassed a variety of configurations, including sampling, quantization, signal normalization, silence removal, Wiener filtering, data scaling, windowing, augmentation, and classifier tuning using XGBoost. Through the analysis of time, frequency, mel-frequency, and statistical features, we achieved an impressive accuracy of 99.54% and an F-Beta score of 99.52% with just 6 boosting trees at an 8 kHz, 8-bit configuration. Moreover, when utilizing only MFCCs along with their first- and second-order deltas, we recorded an accuracy of 97.83% and an F-Beta score of 97.67%. Lastly, by implementing a greedy wrapper approach, we obtained a remarkable accuracy of 96.82% and an F-Beta score of 98.86% using 50 selected features, nearly all of which were first- and second-order deltas of the MFCCs.
△ Less
Submitted 24 November, 2024; v1 submitted 9 November, 2024;
originally announced November 2024.
-
Simultaneous optimization of control gains and reference filter coefficients for trajectory tracking control
Authors:
Amane Sakanashi,
Rin Suyama,
Atsuo Maki,
Youhei Akimoto
Abstract:
Research on vessel automation and autonomy is currently being conducted by various countries and institutions. Safe and accurate ship control algorithms are crucial to realize automated operation. Actuator drive constraints of a target ship may jeopardize the stability of the control law and require complex theory. In this study, we include a penalty term to the control law gain optimization stage…
▽ More
Research on vessel automation and autonomy is currently being conducted by various countries and institutions. Safe and accurate ship control algorithms are crucial to realize automated operation. Actuator drive constraints of a target ship may jeopardize the stability of the control law and require complex theory. In this study, we include a penalty term to the control law gain optimization stage of dynamic positioning systems to account for the amounts by which the actuator input value and its rate of change exceed the constraint. The parameters for generating a suitable reference path for the control law are identified simultaneously with the control gains. The simulation results show that the proposed method can realize control parameters and a reference design with excellent tracking performance while determining the cost of the controller design by considering the effects of both the actuators and rate saturation.
△ Less
Submitted 8 July, 2024;
originally announced July 2024.
-
A Practical and Online Trajectory Planner for Autonomous Ships' Berthing, Incorporating Speed Control
Authors:
Agnes Ngina Mwange,
Dimas Maulana Rachman,
Rin Suyama,
Atsuo Maki
Abstract:
Autonomous ships are essentially designed and equipped to perceive their internal and external environment and subsequently perform appropriate actions depending on the predetermined objective(s) without human intervention. Consequently, trajectory planning algorithms for autonomous berthing must consider factors such as system dynamics, ship actuators, environmental disturbances, and the safety o…
▽ More
Autonomous ships are essentially designed and equipped to perceive their internal and external environment and subsequently perform appropriate actions depending on the predetermined objective(s) without human intervention. Consequently, trajectory planning algorithms for autonomous berthing must consider factors such as system dynamics, ship actuators, environmental disturbances, and the safety of the ship, other ships, and port structures, among others. In this study, basing the ship dynamics on the low-speed MMG model, trajectory planning for an autonomous ship is modeled as an optimal control problem (OCP) that is transcribed into a nonlinear programming problem (NLP) using the direct multiple shooting technique. To enhance berthing safety, besides considering wind disturbances, speed control, actuators' limitations, and collision avoidance features are incorporated as constraints in the NLP, which is then solved using the Sequential Quadratic Programming (SQP) algorithm in MATLAB. Finally, the performance of the proposed planner is evaluated through (i) comparison with solutions obtained using CMA-ES for two different model ships, (ii) trajectory planning for different harbor entry and berth approach scenarios, and (iii) feasibility study using stochastically generated initial conditions and positions within the port boundaries. Simulation results indicate enhanced berthing safety as well as practical and computational feasibility making the planner suitable for real-time applications.
△ Less
Submitted 14 February, 2024;
originally announced February 2024.
-
Parameter fine-tuning method for MMG model using real-scale ship data
Authors:
Rin Suyama,
Rintaro Matsushita,
Ryo Kakuta,
Kouki Wakita,
Atsuo Maki
Abstract:
In this paper, a fine-tuning method of the parameters in the MMG model for the real-scale ship is proposed. In the proposed method, all of the arbitrarily indicated target parameters of the MMG model are tuned simultaneously in the framework of SI using time series data of real-sale ship maneuvering motion data to steadily improve the accuracy of the MMG model. Parameter tuning is formulated as a…
▽ More
In this paper, a fine-tuning method of the parameters in the MMG model for the real-scale ship is proposed. In the proposed method, all of the arbitrarily indicated target parameters of the MMG model are tuned simultaneously in the framework of SI using time series data of real-sale ship maneuvering motion data to steadily improve the accuracy of the MMG model. Parameter tuning is formulated as a minimization problem of the deviation of the maneuvering motion simulated with given parameters and the real-scale ship trials, and the global solution is explored using CMA-ES. By constraining the exploration ranges to the neighborhood of the previously determined parameter values, the proposed method limits the output in a realistic range. The proposed method is applied to the tuning of 12 parameters for a container ship with five different widths of the exploration range. The results show that, in all cases, the accuracy of the maneuvering simulation is improved by applying the tuned parameters to the MMG model, and the validity of the proposed parameter fine-tuning method is confirmed.
△ Less
Submitted 7 December, 2023;
originally announced December 2023.
-
Nonlinear steering control under input magnitude and rate constraints with exponential convergence
Authors:
Rin Suyama,
Satoshi Satoh,
Atsuo Maki
Abstract:
A ship steering control is designed for a nonlinear maneuvering model whose rudder manipulation is constrained in both magnitude and rate. In our method, the tracking problem of the target heading angle with input constraints is converted into the tracking problem for a strict-feedback system without any input constraints. To derive this system, hyperbolic tangent ($\tanh$) function and auxiliary…
▽ More
A ship steering control is designed for a nonlinear maneuvering model whose rudder manipulation is constrained in both magnitude and rate. In our method, the tracking problem of the target heading angle with input constraints is converted into the tracking problem for a strict-feedback system without any input constraints. To derive this system, hyperbolic tangent ($\tanh$) function and auxiliary variables are introduced to deal with the input constraints. Furthermore, using the feature of the derivative of $\tanh$ function, auxiliary systems are successfully derived in the strict-feedback form. The backstepping method is utilized to construct the feedback control law for the resulting cascade system. The proposed steering control is verified in numerical experiments, and the result shows that the tracking of the target heading angle is successful using the proposed control law.
△ Less
Submitted 25 October, 2023;
originally announced October 2023.
-
Microphone Array Based Surveillance Audio Classification
Authors:
Dimitri Leandro de Oliveira Silva,
Tito Spadini,
Ricardo Suyama
Abstract:
The work assessed seven classical classifiers and two beamforming algorithms for detecting surveillance sound events. The tests included the use of AWGN with -10 dB to 30 dB SNR. Data Augmentation was also employed to improve algorithms' performance. The results showed that the combination of SVM and Delay-and-Sum (DaS) scored the best accuracy (up to 86.0\%), but had high computational cost (…
▽ More
The work assessed seven classical classifiers and two beamforming algorithms for detecting surveillance sound events. The tests included the use of AWGN with -10 dB to 30 dB SNR. Data Augmentation was also employed to improve algorithms' performance. The results showed that the combination of SVM and Delay-and-Sum (DaS) scored the best accuracy (up to 86.0\%), but had high computational cost ($\approx $ 402 ms), mainly due to DaS. The use of SGD also seems to be a good alternative since it has achieved good accuracy either (up to 85.3\%), but with quicker processing time ($\approx$ 165 ms).
△ Less
Submitted 22 May, 2020;
originally announced May 2020.
-
Sound Event Recognition in a Smart City Surveillance Context
Authors:
Tito Spadini,
Dimitri Leandro de Oliveira Silva,
Ricardo Suyama
Abstract:
Due to the growing demand for improving surveillance capabilities in smart cities, systems need to be developed to provide better monitoring capabilities to competent authorities, agencies responsible for strategic resource management, and emergency call centers. This work assumes that, as a complementary monitoring solution, the use of a system capable of detecting the occurrence of sound events,…
▽ More
Due to the growing demand for improving surveillance capabilities in smart cities, systems need to be developed to provide better monitoring capabilities to competent authorities, agencies responsible for strategic resource management, and emergency call centers. This work assumes that, as a complementary monitoring solution, the use of a system capable of detecting the occurrence of sound events, performing the Sound Events Recognition (SER) task, is highly convenient. In order to contribute to the classification of such events, this paper explored several classifiers over the SESA dataset, composed of audios of three hazard classes (gunshots, explosions, and sirens) and a class of casual sounds that could be misinterpreted as some of the other sounds. The best result was obtained by SGD, with an accuracy of 72.13% with 6.81 ms classification time, reinforcing the viability of such an approach.
△ Less
Submitted 1 February, 2020; v1 submitted 27 October, 2019;
originally announced October 2019.
-
Comparative Study between Adversarial Networks and Classical Techniques for Speech Enhancement
Authors:
Tito Spadini,
Ricardo Suyama
Abstract:
Speech enhancement is a crucial task for several applications. Among the most explored techniques are the Wiener filter and the LogMMSE, but approaches exploring deep learning adapted to this task, such as SEGAN, have presented relevant results. This study compared the performance of the mentioned techniques in 85 noise conditions regarding quality, intelligibility, and distortion; and concluded t…
▽ More
Speech enhancement is a crucial task for several applications. Among the most explored techniques are the Wiener filter and the LogMMSE, but approaches exploring deep learning adapted to this task, such as SEGAN, have presented relevant results. This study compared the performance of the mentioned techniques in 85 noise conditions regarding quality, intelligibility, and distortion; and concluded that classical techniques continue to exhibit superior results for most scenarios, but, in severe noise scenarios, SEGAN performed better and with lower variance.
△ Less
Submitted 21 October, 2019;
originally announced October 2019.
