CSE
CSE
CSE CSE


Computer Engineering Seminar

Data Storage and Querying for Resource Constrained Computing Devices

Graeme Douglas


 
Thursday, October 22, 2015
12:00pm - 1:30pm
3725 BBB

Add to Google Calendar

About the Event

Database management systems (DBMSs) have greatly reduced the costs associated with collecting, storing, and making insights from data. Increasingly, small and energy efficient devices are being used to collect and manage important data. Applications of ubiquitous computing such as building control, smart traffic management, and home automation all require reliable data management solutions for inexpensive computing platforms. This work implements data storage and query techniques used towards producing reliable database systems for highly resource constrained devices. In particular, LittleD provides the first SQL language relational database using 2KB of memory or less for most queries. Additionally, IonDB is a key-value store for embedded systems that provides a simple but powerful API for record storage, iteration, and retrieval.

Biography

Graeme earned his B. Sc. in Computer Science (honours) and Mathematics (honours) with from The University of British Columbia’s Okanagan campus in 2015. During his undergraduate studies, Graeme developed and published research on new data management technologies for extremely resource constrained devices. In particular, he is the sole creator of LittleD, the only relational database processing SQL queries in 2KB of RAM or less. Additionally, Graeme is a key architect and contributor to IonDB, a key-value store for Arduino’s and other embedded systems. In industry, Graeme has worked with companies both small and large on difficult problems such as automated decision making under uncertainty and database performance optimization.

Additional Information

Contact: Stephen Reger

Phone: 734-764-9401

Email: sereger@eecs.umich.edu

Sponsor(s): CE Lab

Faculty Sponsor: Prabal Dutta

Open to: Public