Thoughts on Edge Sensing and Processing

The challenges of edge sensing and processing in commercial and military domains are reviewed and approaches for addressing them identified. Current developments and limitations in integrating mobile devices with cloud capabilities are presented and potential for cloudlets and fog computing to fill the gaps is described. Growth in requirements is expected in nearly every area that is or will be touched by machine learning, including multi-modal image and activity recognition, large language models, and even the ultimate edge, processing in space. Innovative approaches and algorithms will be required to implement these new requirements at the edge.