A Multi Agent System architecture to implement Collaborative Learning for social industrial assets

Bakliwal, Kshitij; Dhada, Maharshi; Palau, Adrià Salvador; Parlikad, Ajith Kumar; Lad, Bhupesh Kumar

doi:10.1016/j.ifacol.2018.08.421

Cited by 31 publications

(21 citation statements)

References 13 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The architectures reviewed in this paper are formed by four elements: Virtual Assets, Digital Twins, Mediator Agents, and a Social Platform. Some of these agents were already described in several publications (Palau et al 2019a;Bakliwal et al 2018;Palau et al 2018Palau et al , 2019b, and their description here is inspired in the original papers. The agent's failure modes have been restricted to affect their deliberative and communicative capabilities.…”

Section: Agent Typologies and Their Failure Modesmentioning

confidence: 99%

Multi-agent system architectures for collaborative prognostics

2019

Self Cite

View full text Add to dashboard Cite

This paper provides a methodology to assess the optimal multi-agent architecture for collaborative prognostics in modern fleets of assets. The use of multi-agent systems has been shown to improve the ability to predict equipment failures by enabling machines with communication and collaborative learning capabilities. Different architectures have been postulated for industrial multi-agent systems in general. A rigorous analysis of the implications of their implementation for collaborative prognostics is essential to guide industrial deployment. In this paper, we investigate the cost and reliability implications of using different multi-agent systems architectures for collaborative failure prediction and maintenance optimization in large fleets of industrial assets. Results show that purely distributed architectures are optimal for high-value assets, while hierarchical architectures optimize communication costs for low-value assets. This enables asset managers to design and implement multi-agent systems for predictive maintenance that significantly decrease the whole-life cost of their assets.

show abstract

Section: Agent Typologies and Their Failure Modesmentioning

confidence: 99%

Multi-agent system architectures for collaborative prognostics

2019

Self Cite

View full text Add to dashboard Cite

show abstract

“…The architecture used in this paper is a modified hierarchical architecture with three layers: Virtual Assets, Digital Twins, and a Social Platform (see Figure 1) [9], [14]. Such kind of architecture is a modification of a purely hierarchical architecture, featuring also horizontal communication between agents on the same level [9].…”

Section: System Description a Architecturementioning

confidence: 99%

“…In the sections that follow, we describe how we have designed the Virtual Assets, Digital Twins and Social Platform in order to adapt them to perform distributed collaborative prognostics. The original architecture, designed to facilitate collaborative learning, is described in more detail in [14]. 1) Virtual Asset: as the lowest-level blocks of the system's architecture, the Virtual Assets are originally designed to standardize the data coming from their corresponding assets.…”

Section: System Description a Architecturementioning

confidence: 99%

“…1) Virtual Asset: as the lowest-level blocks of the system's architecture, the Virtual Assets are originally designed to standardize the data coming from their corresponding assets. However, another function from these architectural elements can be to simulate machines working in real time from an existing prognostics data set [14]. In the later case, the Virtual Assets load the data from a data file, convert it into the system's standard format and send it at fixed time intervals to their assigned Digital Twins.…”

Section: System Description a Architecturementioning

confidence: 99%

“…In this paper, we present the first implementation of distributed collaborative prognostics using a combination of stateof-the-art prognostics algorithms and an ad-hoc architecture extracted from [14]. In the proposed implementation, each asset is assigned a Digital Twin that processes asset data and performs prognostics.…”

Section: Introductionmentioning

confidence: 99%

See 2 more Smart Citations

Recurrent Neural Networks for real-time distributed collaborative prognostics

Palau

Bakliwal

Dhada

et al. 2018

2018 IEEE International Conference on Prognostics and Health Management (ICPHM)

Self Cite

View full text Add to dashboard Cite

We present the first steps towards real-time distributed collaborative prognostics enabled by an implementation of the Weibull Time To Event-Recurrent Neural Network (WTTE-RNN) algorithm. In our system, assets determine their time to failure (TTF) in real-time according to an asset-specific model that is obtained in collaboration with other similar assets in the asset fleet. The presented approach builds on the emergent field of similarity analysis in asset management, and extends it to distributed collaborative prognostics. We show how through collaboration between assets and distributed prognostics, competitive time to failure estimates can be obtained. 1

show abstract

Multi Agent System for Machine Learning Under Uncertainty in Cyber Physical Manufacturing System

Yong

Brintrup

2019

Service Oriented, Holonic and Multi-Agent Manufacturing Systems for Industry of the Future

View full text Add to dashboard Cite

Recent advancements in predictive machine learning has led to its application in various use cases in manufacturing. Most research focused on maximising predictive accuracy without addressing the uncertainty associated with it. While accuracy is important, focusing primarily on it poses an overfitting danger, exposing manufacturers to risk, ultimately hindering the adoption of these techniques. In this paper, we determine the sources of uncertainty in machine learning and establish the success criteria of a machine learning system to function well under uncertainty in a cyber-physical manufacturing system (CPMS) scenario. Then, we propose a multi-agent system architecture which leverages probabilistic machine learning as a means of achieving such criteria. We propose possible scenarios for which our proposed architecture is useful and discuss future work. Experimentally, we implement Bayesian Neural Networks for multi-tasks classification on a public dataset for the real-time condition monitoring of a hydraulic system and demonstrate the usefulness of the system by evaluating the probability of a prediction being accurate given its uncertainty. We deploy these models using our proposed agent-based framework and integrate web visualisation to demonstrate its real-time feasibility.

show abstract

A Multi Agent System architecture to implement Collaborative Learning for social industrial assets

Cited by 31 publications

References 13 publications

Multi-agent system architectures for collaborative prognostics

Multi-agent system architectures for collaborative prognostics

Recurrent Neural Networks for real-time distributed collaborative prognostics

Multi Agent System for Machine Learning Under Uncertainty in Cyber Physical Manufacturing System

Contact Info

Product

Resources

About