536: I’ve Seen a Giraffe with David Such
Catch up with our latest episode and embedded-related news from around The Web…
536: I’ve Seen a Giraffe
David Such joins us to discuss his new book, Embedded AI: Intelligence at the Deep Edge. We explore the boundary between classical signal processing, heuristics, and edge machine learning; IMU processing and sensor fusion; and how to determine what sort of ML fits on your processor. David also shares his experiences writing for No Starch Press and why future robotic architectures need a digital cerebellum for low-level reflexes.
David Such is an electrical engineer based in Sydney. He is the director of Reefwing Software.
DISCUSSION RESOURCES & LINKS
David's Website & Book Resources:
Embedded AI: Intelligence at the Deep Edge (No Starch Press, on Amazon)
Technologies & Frameworks Mentioned:
Sensor Fusion, Kalman filters, Mahoney filters, and quaternions for IMU processing
Raspberry Pi Pico / RP2040 PIO and DMA engines for high-speed signal processing
Syntiant neural processing chips (Arduino Nicla Voice)
MicroPython for embedded scripting and test automation
Book Giveaway & Discount Code:
No Starch Press is giving away two print copies of Embedded AI (US/Canada residents). Entry rules are in the show, due by Nov 1, 2026. There is also a coupon for the book good through Nov 5, 2026.
Check out the full transcript, and feel free to share, comment on, and post your favorite excerpt.
Episode Excerpt
David (53:31):
So I would argue that we need to go to a different sort of model, where the safety is built in lower down. Same way that the evolution has worked it out. We have reflexes that do not go to the neocortex until after they have already fired. So I think that is the sort of model that we need to move to, in terms of robotics and self-driving cars and things like that.
Elecia (53:58):
It is funny you brought up MicroPython. To some extent, one of the reasons I really liked it was because I could build in all of those safety [checks] in the C, so they could not get to anything that was going to blow up my robot. That was super helpful.
Your neocortex versus I guess, amygdala or cortex or I do not know, however many brain parts you want. It does make sense to separate it into, "Here are the safety critical pieces." And if you hear the screech of the cat's claws on a rock, which sounds a lot like the screech of a blackboard, yes, your brain is going to go, "Wow. We need to do things."
Chris (54:41):
That is a good principle even without AI. There are plenty of embedded systems I have worked on, where we put the safety systems in hardware. "Oh, the laser is firing? Well, there is a hardware integrator that measures how much energy is put in, and if it goes too much, it shuts it off."
Instead of the software, "Oh, I am looping around sampling. It seems all right. Oops, I got distracted and the laser was on too long."
Elecia (55:03):
I love that firmware engineers are the people who least trust the firmware.
Some links we’ve found interesting:
The Embedded Online Conference call for speakers is open until November 8. If you’ve got something you want to cover, now’s the time to put together a proposal.
Hasbro has opened submissions for its fourth annual Women Innovators of Play Design Challenge. It’s a contest to come up with an original concept for a new toy or game. Submissions are open until November 20, 2026 and we’ll talk about it on the upcoming show.
Embedded Swift continues to evolve with several new language features. Chris is looking into trying this out in the next few weeks. We’re curious whether anyone out there is doing Embedded Swift work.
Shawn Hymel has started a new video series on Reinforcement Learning for Robotics over at Digikey. The example project will be a self-balancing two-wheeled robot. Shawn was on episode 403.
Upcoming Show
This Thursday, it’ll be just Elecia and Chris, and we’ll be talking about daydreaming about toys, follow-up from our various LLM-focused episodes, and more.
Patreon and Ko-fi
To join the Embedded conversation on Slack, support us on Patreon or Ko-fi. Members at the $5-a-month level receive an ad-free feed and special episodes that are not available to the public.
Interested in sponsoring the newsletter or the show? Drop us a line at sponsorship at embedded.fm and let's talk! ☎️
Was this email forwarded to you? Sign up here.
Embedded.fm, 500 Cathedral Dr #2301, Aptos, CA 95003, United States of America
Powered by Squarespace