Abstract: While Python and R dominate day to day tasks for typical analytics and prediction in both batch and real-time processing in the cloud, a need for moving closer to the source of the data has arisen. The Smart Cities Initiative begun under the Obama administration, as one of its key strategies, seeks to turn the US into a global leader for “Internet of Things,” a ubiquitous network of connected devices, smart sensors, and big data analytics.
However, rather than just choke off bandwidth passing every item to the cloud, Fog Computing has surfaced as a means to decentralize the computing infrastructure in which data, compute, storage and applications are distributed in the most logical, efficient place between the data source and the cloud. But where will the processing occur?
Fog computing provides the concept of data and compute planes, and in this talk, Dr. Maskrey will discuss a representative implantation she is working on utilizing new features of Apple’s iOS 11 to move predictive models into the cloud and closer to the point of origin of the date. New frameworks and tools allow for the utilization of existing, trained models down at the mobile device level. For example, a single secure iPad device, functioning as a secure IoT aggregator point for various IoT sensor, can implement, directly, a convolutional neural network for image process to, for example, identify empty parking spaces. The model used can be a standard Keras, Caffe, or even scikit-learn model of almost any form.
In just the past decade, the convenience tools we once used from our work or home computer all now reside on our phone and even our watch. The next decade will see a similar transition from cloud services where big data analytic process will continue to function in the cloud, but where smaller data, like that of a parking lot or neighborhood water monitoring system will move to the fog.
Bio: Dr. Maskrey obtained her PhD in applied mathematics and her MSEE with a specialization in communications theory focusing on non-linear techniques. In addition to operating her two companies, having completed a successful exit of a third late 2016, Molly consults in the Denver, Colorado area on topics ranging from brain imagery/perfusion analysis for cognitive disorders, early childhood development, gait and movement analysis for Cerebral Palsy patients, and CTIA over the air testing of mobile devices. Dr. Maskrey has written five technical books on mobile technology in the past two years.
Dr. Molly Maskrey
Data Scientist/Electrical Engineer/iOS-Swift Author