TinyMLOps : operational challenges for widespread edge AI adoption
- Author
- Sam Leroux (UGent) , Pieter Simoens (UGent) , Meelis Lootus, Kartik Thakore and Akshay Sharma
- Organization
- Abstract
- Deploying machine learning applications on edge devices can bring clear benefits such as improved reliability, latency and privacy but it also introduces its own set of challenges. Most works focus on the limited computational resources of edge platforms but this is not the only bottleneck standing in the way of widespread adoption. In this paper we list several other challenges that a TinyML practitioner might need to consider when operationalizing an application on edge devices. We focus on tasks such as monitoring and managing the application, common functionality for a MLOps platform, and show how they are complicated by the distributed nature of edge deployment. We also discuss issues that are unique to edge applications such as protecting a model's intellectual property and verifying its integrity.
- Keywords
- DEEP NEURAL-NETWORKS, INFERENCE, INTERNET, TinyML, Edge AI, MLOps, TinyMLOps
Downloads
-
(...).pdf
- full text (Published version)
- |
- UGent only
- |
- |
- 187.37 KB
Citation
Please use this url to cite or link to this publication: http://hdl.handle.net/1854/LU-8768562
- MLA
- Leroux, Sam, et al. “TinyMLOps : Operational Challenges for Widespread Edge AI Adoption.” 2022 IEEE 36TH INTERNATIONAL PARALLEL AND DISTRIBUTED PROCESSING SYMPOSIUM WORKSHOPS (IPDPSW 2022), IEEE, 2022, pp. 1003–10, doi:10.1109/IPDPSW55747.2022.00160.
- APA
- Leroux, S., Simoens, P., Lootus, M., Thakore, K., & Sharma, A. (2022). TinyMLOps : operational challenges for widespread edge AI adoption. 2022 IEEE 36TH INTERNATIONAL PARALLEL AND DISTRIBUTED PROCESSING SYMPOSIUM WORKSHOPS (IPDPSW 2022), 1003–1010. https://doi.org/10.1109/IPDPSW55747.2022.00160
- Chicago author-date
- Leroux, Sam, Pieter Simoens, Meelis Lootus, Kartik Thakore, and Akshay Sharma. 2022. “TinyMLOps : Operational Challenges for Widespread Edge AI Adoption.” In 2022 IEEE 36TH INTERNATIONAL PARALLEL AND DISTRIBUTED PROCESSING SYMPOSIUM WORKSHOPS (IPDPSW 2022), 1003–10. IEEE. https://doi.org/10.1109/IPDPSW55747.2022.00160.
- Chicago author-date (all authors)
- Leroux, Sam, Pieter Simoens, Meelis Lootus, Kartik Thakore, and Akshay Sharma. 2022. “TinyMLOps : Operational Challenges for Widespread Edge AI Adoption.” In 2022 IEEE 36TH INTERNATIONAL PARALLEL AND DISTRIBUTED PROCESSING SYMPOSIUM WORKSHOPS (IPDPSW 2022), 1003–1010. IEEE. doi:10.1109/IPDPSW55747.2022.00160.
- Vancouver
- 1.Leroux S, Simoens P, Lootus M, Thakore K, Sharma A. TinyMLOps : operational challenges for widespread edge AI adoption. In: 2022 IEEE 36TH INTERNATIONAL PARALLEL AND DISTRIBUTED PROCESSING SYMPOSIUM WORKSHOPS (IPDPSW 2022). IEEE; 2022. p. 1003–10.
- IEEE
- [1]S. Leroux, P. Simoens, M. Lootus, K. Thakore, and A. Sharma, “TinyMLOps : operational challenges for widespread edge AI adoption,” in 2022 IEEE 36TH INTERNATIONAL PARALLEL AND DISTRIBUTED PROCESSING SYMPOSIUM WORKSHOPS (IPDPSW 2022), Lyon, France, 2022, pp. 1003–1010.
@inproceedings{8768562,
abstract = {{Deploying machine learning applications on edge devices can bring clear benefits such as improved reliability, latency and privacy but it also introduces its own set of challenges. Most works focus on the limited computational resources of edge platforms but this is not the only bottleneck standing in the way of widespread adoption. In this paper we list several other challenges that a TinyML practitioner might need to consider when operationalizing an application on edge devices. We focus on tasks such as monitoring and managing the application, common functionality for a MLOps platform, and show how they are complicated by the distributed nature of edge deployment. We also discuss issues that are unique to edge applications such as protecting a model's intellectual property and verifying its integrity.}},
author = {{Leroux, Sam and Simoens, Pieter and Lootus, Meelis and Thakore, Kartik and Sharma, Akshay}},
booktitle = {{2022 IEEE 36TH INTERNATIONAL PARALLEL AND DISTRIBUTED PROCESSING SYMPOSIUM WORKSHOPS (IPDPSW 2022)}},
isbn = {{9781665497473}},
issn = {{2164-7062}},
keywords = {{DEEP NEURAL-NETWORKS,INFERENCE,INTERNET,TinyML,Edge AI,MLOps,TinyMLOps}},
language = {{eng}},
location = {{Lyon, France}},
pages = {{1003--1010}},
publisher = {{IEEE}},
title = {{TinyMLOps : operational challenges for widespread edge AI adoption}},
url = {{http://doi.org/10.1109/IPDPSW55747.2022.00160}},
year = {{2022}},
}
- Altmetric
- View in Altmetric
- Web of Science
- Times cited: