TinyML for Edge Networks: Challenges and Future Directions

TinyML, deploying low-complexity Machine Learning (ML) models on microcontroller units, which have constrained the availability of resources, is revolutionizing the technologies for extremely low-end devices. This study investigates the importance, categorization, and uses of TinyML in edge networks, providing an overview of workflow and the importance of different compression techniques. TinyML can be applied in various industries such as healthcare, Industrial Internet of Things (IIoT), and agriculture where it demonstrates the capability of doing local data processing and fast decision making which means need to provide the results there on spot. Nevertheless, TinyML encounters limitations and challenges such as interoperability, memory usage, energy efficiency, and standardization, which are also discussed in this article.