Federated Learning for Advanced Manufacturing Based on Industrial IoT Data Analytics

Kanagavelu, Renuga; Li, Zengxiang; Samsudin, Juniarto; Hussain, Shaista; Yang, Feng; Yang, Yechao; Goh, Rick Siow Mong; Cheah, Mervyn

doi:10.1007/978-3-030-67270-6_6

Cited by 8 publications

(7 citation statements)

References 10 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…and such hash values with corresponding sample indexes and labels are uploaded to the server for initializing the knowledge cache according to Eq. (7,8,9,10,11).…”

Section: Knowledge Cache-driven Personalized Distillationmentioning

confidence: 99%

“…Update LI, IK, IH according to Eq. (7,8,9) 5 until Receive all indexes (k, i) from K clients; 6 Build relations via HNSW [43] according to Eq. (10) and Eq.…”

Section: Formal Description Of Fedcachementioning

confidence: 99%

“…Federated Learning (FL) is a privacy-preserving distributed learning paradigm that enables multiple data owners to collaboratively train AI models without sharing owners' private data. Due to the benefits of data localization and privacy protection, FL has shown great potential in various EI applications, such as healthcare [5], smart transportation [6], industrial manufacturing [7], etc. Unfortunately, the prevailing FL approaches [8], [9] require all participating devices (named clients) to share a uniform model, which is extremely difficult to deploy and generalize to all devices because of the inherent characteristics of device variation in terms of data heterogeneity, resources limitation, task differentiation, etc [10], [11].…”

mentioning

confidence: 99%

See 2 more Smart Citations

FedCache: A Knowledge Cache-driven Federated Learning Architecture for Personalized Edge Intelligence

Wu,

Sun,

Wang

et al. 2024

Preprint

View full text Add to dashboard Cite

Edge Intelligence (EI) enables Artificial Intelligence (AI) applications to run at the edge, where data analysis and decision-making can be performed in real-time and close to data sources. To protect data privacy and unify data silos distributed among end devices in EI, Federated Learning (FL) is proposed for collaborative training shared AI models across multiple devices without compromising data security. However, the prevailing FL approaches cannot guarantee model generalization and adaptation on heterogeneous clients. Recently, Personalized Federated Learning (PFL) has drawn growing awareness in EI, as it enables striking a productive balance between local-specific training requirements inherent in devices and global-generalized optimization objectives for satisfactory performance. However, most existing PFL methods are based on the Parameters Interaction-based Architecture (PIA) represented by FedAvg, which causes unaffordable communication burdens due to large-scale parameters transmission between devices and the edge server. In contrast, Logits Interaction-based Architecture (LIA) enables to update model parameters with logits transfer, and gains the advantages of communication lightweight and heterogeneous on-device model allowance compared to PIA. Nevertheless, previous LIA methods attempt to achieve satisfactory performance either relying on unrealistic public datasets or increasing communication overhead for additional information transmission other than logits. To tackle this dilemma, we propose a knowledge cache-driven PFL architecture, named FedCache, which reserves a knowledge cache on the server for fetching personalized knowledge from the samples with similar hashes to each given on-device sample. During the training phase, ensemble distillation is applied to on-device models for constructive optimization with personalized knowledge transferred from the server-side knowledge cache. Empirical experiments on four datasets demonstrate the comparable performance of FedCache with state-of-art PFL approaches, with more than two orders of magnitude improvements in communication efficiency. Our code and DEMO are available at https://github.com/wuzhiyuan2000/FedCache.

show abstract

“…and such hash values with corresponding sample indexes and labels are uploaded to the server for initializing the knowledge cache according to Eq. (7,8,9,10,11).…”

Section: Knowledge Cache-driven Personalized Distillationmentioning

confidence: 99%

“…Update LI, IK, IH according to Eq. (7,8,9) 5 until Receive all indexes (k, i) from K clients; 6 Build relations via HNSW [43] according to Eq. (10) and Eq.…”

Section: Formal Description Of Fedcachementioning

confidence: 99%

mentioning

confidence: 99%

See 1 more Smart Citation

FedCache: A Knowledge Cache-driven Federated Learning Architecture for Personalized Edge Intelligence

Wu,

Sun,

Wang

et al. 2024

Preprint

View full text Add to dashboard Cite

show abstract

“…Numerous modern technologies have been used as computational backbones for collecting user data and how to share it with others, services such as cloud computing, radio-frequency identification, and security solutions to safeguard the user data against hackers [129]. Federated learning [130] enables the mining of data that is dispersed across multiple sites.…”

Section: ) Digital Data Managementmentioning

confidence: 99%

Blockchain for Industry 5.0: Vision, Opportunities, Key Enablers, and Future Directions

et al. 2022

View full text Add to dashboard Cite

Industry 4.0 have witnessed a paradigm shift from cyber-physical systems (CPS) that aims at massive automation, towards a more customer-driven approach. The shift has been attributed to the design of hyper-cognitive systems, integration of virtual and extended reality, digital machinery prototyping and twin designs, trusted machine boundaries, collaborative robots, and artificial intelligence (AI)-based supply chains. This new wave, termed Industry 5.0, is expected to leverage massive production with user-centric customization outside the scope of Industry 4.0 ecosystems. Industry 5.0 is expected to assist diverse industrial verticals like healthcare, smart farming, drones, smart grids, and supply chain production ecosystems. However, data is shared among multiple heterogeneous networks, spanning different authoritative domains. Thus, trusted and secured data transfer is crucial to synergize and secure the industrial perimeters. Blockchain (BC) is a preferred choice as a security enabler to Industry 5.0 ecosystems owing to its inherent property of immutability, chronology, and auditability in industrial systems. Limited works are proposed that present the vision and holistic view of BC-assisted Industry 5.0 applications. The article presents a first-of-its-kind survey on BC as a security enabler in Industry 5.0. Based on a descriptive survey methodology and research questions, we presented the key drivers, and potential applications, and propose an architectural vision of BC-based Industry 5.0 in diverse applicative verticals. The survey intends to present solutions that would assist industry practitioners, academicians, and researchers to drive novel BCassisted solutions in Industry 5.0 verticals.

show abstract

“…In manufacturing industries, FL enable multiple companies to train a condition-monitoring system, that monitors a particular condition in machinery (such as vibration, temperature, etc.) to identify changes that could indicate a developing fault, without revealing their respective data and assets [23] and keep their IP confidential. However, despite this growing popularity [45], FL already has been challenged by several privacy attacks, in particular because the shared models can be reverse engineered to identify clients data, or at least some of its features.…”

Section: Federated Learning (Fl)mentioning

confidence: 99%

Shielding Federated Learning Systems against Inference Attacks with ARM TrustZone

Messaoud¹,

Mokhtar²,

Nitu³

et al. 2022

Preprint

View full text Add to dashboard Cite

Federated Learning (FL) opens new perspectives for training machine learning models while keeping personal data on the users premises. Specifically, in FL, models are trained on the users' devices and only model updates (i.e., gradients) are sent to a central server for aggregation purposes. However, the long list of inference attacks that leak private data from gradients, published in the recent years, have emphasized the need of devising effective protection mechanisms to incentivize the adoption of FL at scale. While there exist solutions to mitigate these attacks on the server side, little has been done to protect users from attacks performed on the client side. In this context, the use of Trusted Execution Environments (TEEs) on the client side are among the most proposing solutions. However, existing frameworks (e.g., DarkneTZ) require statically putting a large portion of the machine learning model into the TEE to effectively protect against complex attacks or a combination of attacks. We present GradSec, a solution that allows protecting in a TEE only sensitive layers of a machine learning model, either statically or dynamically, hence reducing both the TCB size and the overall training time by up to 30% and 56%, respectively compared to state-of-the-art competitors.

show abstract

Federated Learning for Advanced Manufacturing Based on Industrial IoT Data Analytics

Cited by 8 publications

References 10 publications

FedCache: A Knowledge Cache-driven Federated Learning Architecture for Personalized Edge Intelligence

FedCache: A Knowledge Cache-driven Federated Learning Architecture for Personalized Edge Intelligence

Blockchain for Industry 5.0: Vision, Opportunities, Key Enablers, and Future Directions

Shielding Federated Learning Systems against Inference Attacks with ARM TrustZone

Contact Info

Product

Resources

About