Hardware, Software & the Internet of Things
June 23–25, 2015 • San Francisco, CA

Intro to inertial sensors: From taps to gestures to location

Elecia White (Logical Elegance and Embedded.fm)
2:05pm–2:45pm Thursday, 06/25/2015
Location: Southside Theater (Bldg D)
Average rating: *****
(5.00, 4 ratings)
Slides:   1-PPTX 

Prerequisite Knowledge

This is an introductory talk, there are no prerequisites.


What is the difference between an accelerometer, a gyroscope, and a magnetometer? What would you use each for? If you aren’t sure, let me explain.

The entertaining host of the Embedded.fm podcast, Elecia White will explain the differences and each inertial sensor’s best uses, on their own and in combination. She will detail the most common ways to put them together and help you determine which are the best choices for your products.

The talk will discuss how to replace buttons with accelerometers, how that leads to gesture recognition, and why integrating to get location is a more difficult problem than it sounds. While you might not be able to implement a Kalman filter by the end of the session, you will know why it matters.

Photo of Elecia White

Elecia White

Logical Elegance and Embedded.fm

Author of O’Reilly’s Making Embedded Systems and host of the weekly Embedded podcast, Elecia White has worked on DNA scanners, inertial measurement units for airplanes and race cars, toys for preschoolers, a gunshot location system for catching criminals, and assorted other medical and consumer devices. She is the founder of Logical Elegance, an embedded systems consulting company based in San Jose.

Comments on this page are now closed.


Picture of Elecia White
Elecia White
06/25/2015 3:51am PDT

Hi Tom,
I will be touching on the importance of synchronous sampling but not describing mitigation when you can’t control sampling. You can run a Kalman in multiple passes, updating the particular sensor you measure in this timestamp/pass, letting the estimates stand for the other sensors.

Also, I moved a few things so I should have time for questions (or have stories if there aren’t questions).


Picture of Tom Anderson
Tom Anderson
06/23/2015 8:54am PDT

Might be too advanced a topic, and I saw your tweet about having no extra time, but here is a question anyway: In my experience, the time-base jitter in systems such as phones is not as good as what my algorithms would need to operate properly. Is the fix for this category of problems to use timestamps and not assume evenly spaced points in time, or to use dedicated sensors and processors with a good time base, or settle for suboptimal results?