Advances in optimization procedures have brought about a resurgence in artificial neural networks. This renewed focus on hierarchical representation has proven deep learning to be a powerful approach for various data intensiveindustry problems. Even with rapid advances, the application ofdeep learning can still prove challenging. Techniques often rely on large datasets and high performance compute infrastructures. Additionally, the large variety of model architectures and hyperparameters can make fine-tuning difficult. These issues become further exacerbated as new industries, with varied compute environments and data models, seek to leverage these same optimization techniques. This session aims to provide a platform for participantsto openly discuss the integration of deep learning into thevariedenvironmentsof industrial electronics. Welcome topics include applications inrobotics, automation, industrial systems, andlow-power embedded systems and IoT. Work that considers practical issues such as heterogeneous compute environments, fault tolerance, and energy efficiency are also encouraged.