Internet of Things devices, along with the large data volumes that such devices (can potentially) generate, can have a significant impact on our lives, fuelling the development of critical next-generation services and applications in a variety of application domains (e.g., health care, smart grids, finance, disaster management, agriculture, transportation, and water management). Artificial Intelligence technologies, such as Distributed Learning and Training, is finding application in multiple IoT application domains driven by the availability of diverse and large datasets. One such example is the advances in medical diagnostics and prediction that use deep learning technology to improve human health. However, timely and reliable transfer of large data streams (a requirement of deep learning technologies for achieving high accuracy) to centralized locations, such as cloud data centre environments, is being seen as a key limitation of expanding the application horizons of such technologies.

To this end, various paradigms, including osmotic computing, have been proposed that promote the distribution of data analysis tasks across cloud and edge computing environments. However, these existing paradigms fail to provide a detailed account of how technologies such as distributed deep learning can be orchestrated and take advantage of the cloud, edge, and mobile edge environments in a holistic manner. This keynote analyses different algorithmic and programming research challenges involved with the development of holistic and distributed learning algorithms that are resource and data-aware and can account for underlying heterogeneous data models, resource (cloud vs. edge vs. mobile edge) models, and data availability while executing—trading accuracy for execution time, etc.

1. Introduction to the fundamental concepts related to the Osmotic computing paradigm
2. Overview of the research and programming challenges involved with composing and orchestrating complex distributed learning algorithms and workflows in the (cloud-edge) Osmotic computing paradigm
3. Present a novel approach about how to train one Distributed Deep Learning (DDL) model on the hardware of thousands of mid-sized IoT and Edge devices across the world, rather than the use of GPU clusters available within a cloud data centre.
4. Discuss our initial experimental validation using the United Kingdom’s largest IoT infrastructure, namely, the Urban Observatory (http://www.urbanobservatory.ac.uk/)

Professor Rajiv Ranjan is an Australian-British computer scientist, of Indian origin, known for his research in Distributed Systems (Cloud Computing, Big Data, and the Internet of Things). He is University Chair Professor for the Internet of Things research in the School of Computing of Newcastle University, United Kingdom. He is an internationally established scientist in the area of Distributed Systems (having published about 250 scientific papers). He has secured more than $32 Million AUD (£16 Million+ GBP) in the form of competitive research grants from both public and private agencies. He is an innovator with strong and sustained academic and industrial impact and a globally recognized R&D leader with a proven track record. He serves on the editorial boards of top quality international journals including IEEE Transactions on Computers (2014-2016), IEEE Transactions on Cloud Computing, ACM Transactions on the Internet of Things, The Computer (Oxford University), and The Computing (Springer) and Future Generation Computer Systems. He led the Blue Skies section (department, 2014-2019) of IEEE Cloud Computing, where his principal role was to identify and write about the most important, cutting-edge research issues at the intersection of multiple, inter-dependent research disciplines within distributed systems research area including Internet of Things, Big Data Analytics, Cloud Computing, and Edge Computing. He is one of the highly cited authors in computer science and software engineering worldwide (h-index=67, g-index=216, and 23000+ google scholar citations, h-index=35 and 7300+ Scopus citations).