Jonathan Lloyd
Data Scientist - Industrial IoT, Western Digital

Topic:

HDD as an IoT device – Embedded TinyML to enforce policy coordination in storage devices

Abstract:

As storage demands explode and per device capacity continues to grow, the manufacturing test process needs to get more efficient and accurate; this growth involves significant capital expenditure in areas like factory floor space and outlays for test equipment. To address this growing challenge, a form of Narrow AI is gaining traction in HDD manufacturing – an ML model that helps manage specific tasks in the test process. These problems have been typically approached with high end edge or cloud data center resources that communicate with the testing environment.

Recently, we have enforced policy coordination beyond this Narrow AI using the HDD’s own compute resources. We embed low-footprint NN inference capabilities on legacy products without C++11 support, without additional hardware assist or offloading the tasks to testers, host, or network resources.

By utilizing a high volume of resource constrained processors, we demonstrated how an ML-native operation can be robustly used to augment or replace a legacy operation in a mission critical environment.  A tangible increase in backend test throughput on legacy platforms has now been realized.  In this talk we will discuss our case study, lessons learned, best practices developed for this class of application.

Bio:

Jonathan started his lifelong journey with technology at the age of 4 by disassembling and successfully reassembling the family Apple II plus which had the only copy of his father's master’s thesis. Innately curious by nature, he now finds himself a systems design engineer turned data scientist. Specializing in cross-domain embedded computation problems, he is a graduate of Cal Poly Pomona with a MSEE. Prior to joining WDC, he has worked on electric vehicles and lead joint autonomous robot development teams for both Cal Poly Universities. Beginning his tenure at Western Digital, he found ways to create failures from interacting subsystems and created architectural improvements to mitigate them. His work has included primary physics research to develop newly controlled and calibrated manufacturing parameters, authoring many firmware and design changes which enhanced reliability, performance and monitoring capability over the service life of Western Digital products. Unintentionally entering into large scale analytics, he wrote his own ETL pipeline to handle exploratory analysis of field operation phenomena. He is now responsible for design and technical direction of a number of embedded ML activities at WDC.

Please note that in-person conferences (i.e. AI Hardware Expo on May 5th-6th 2020, AI Enterprise Expo on August 25th-26th 2020 and AI Robotics Expo on November 12th-13th 2020 at SEMI, Milpitas, Silicon Valley) are converted into this virtual AI Expo series. 

If you have already purchased a ticket please visit here to access the recordings.