Project Logs

Collapse

• Raspberry Pi 5 Computer Is Functioning Properly After Testing

  Bret Bernhoft • an hour ago • 0 comments

  Previously I had assumed the Raspberry Pi 5 computer I am using for this project was not functioning properly. Thankfully, after updating the operating system, researching possible issues and running a few rudimentary tests, I am pleased to acknowledge the 4GB model seen in the photo below is working as one would expect.

  The two symptoms I observed which led me to assume the RPi 5 was malfunctioning included an inactive cooling fan and unresponsive GPIO pins. It turns out the attached CPU cooling fan only turns on when the computer reaches 50 degrees Celsius. And all the header pins responded to additional sensor testing. So it is full steam ahead into another self-quantification endeavor.

  Ultimately it is important for this Raspberry Pi 5 computer to work correctly because it is central to how this sleep quantification project will be conducted. Without it I will not be able to even launch my experiment.

  Given the confirmation my computer hardware is working, next steps include beginning the assembly of components into a working measurement system. Followed by writing the Python scripts, testing connectivity and ensuring long-term reliability of the whole contraption. Along these lines, I expect the remaining stages of this project to take upwards of 6 months to complete. I am in no rush to finish this experiment and have enjoyed the process thus far.

• Additional Supplies Have Arrived While Others Are In Transit

  Bret Bernhoft • 08/06/2026 at 01:56 • 0 comments

  Another batch of supplies has arrived, including a (1) hot glue gun for relieving strain, (2) adhesive tape for sensor mounting, (3) gauged wire stripper, (4) heat shrink for insulating joints, (5) gaffer tape for cable routing and (6) hardboard panels alongside (7) ribbon cable which will link the sensors back to the Pi.

  Having gathered these supplies, the project has nearly completed its hardware gathering phase. The tools and materials needed to build a setup durable enough for a prolonged study are now in my possession. Remaining steps before soldering begins include finalizing sensor placement on the bed and validating a single sensor channel, while a few final components are still in transit.

• Moving This Project Into Active Development After Hibernation

  Bret Bernhoft • 08/03/2026 at 01:14 • 0 comments

  After a period of hibernation, the Quantifying Sleep Positions With Body Weight project has been restarted. I am approaching this phase with a blank slate mentality, focusing on refining hardware requirements and improving the data pipeline to ensure a widely usable final product. This stage is dedicated to moving past previous hurdles and establishing a stable foundation for long-term sleep pattern analysis.

  Current development is focused on integrating an array of eight FSR pads with a Raspberry Pi 5 to capture pressure data throughout the night. I am currently working on the Python and Flask backend to perfect data collection and storage as JSON. The immediate goal is to establish a proof of concept (POF) capable of providing a reliable baseline for upcoming analyses of physical behavior while asleep.

• Pivoting From Measuring Body Heat To Quantifying Body Weight

  Bret Bernhoft • 03/04/2026 at 01:59 • 0 comments

  Despite having already ordered a new AMG8833 IR thermal sensor, I am choosing to go another route with this project. Instead of quantifying my body heat, I am using 8 FSR pads from Adafruit to measure my body weight. With the same goal of finding patterns within my sleeping positions.

  There remains a tremendous amount to learn from/for this experiment. Pivoting towards different technologies both helps and hinders this process. I am however still optimistic about fruitfully studying my physical resting behaviors while asleep. 

  Onward!

• Attempting To Connect AMG8833 Sensor To Another RPi Computer

  Bret Bernhoft • 02/18/2026 at 03:07 • 0 comments

  Over this past weekend I attempted connecting the AMG8833 IR thermal sensor (being used for this experiment) to another Raspberry Pi computer. This time using a RPi Zero 2 W unit with pre-soldered GPIO header pins.

  As a side note, there were over 150 updates needing to be installed upon first starting up the Raspberry Pi Zero 2 W SBC. Which were completed in a surprisingly short period of time via the CLI.

  Once prepared to tinker, I connected jumper cables to header pins on the thermal camera and then on the Raspberry Pi computer. I next booted the Zero 2 W SBC and ran a simple script to detect hardware interactivity. Unfortunately there was no signal received from the AMG8833 IR thermal camera by the second RPi computer.

  After having two Raspberry Pi computers fail to ping the AMG8833 IR thermal sensor, I assume the problem is the soldering job I did with/on the breakout board's header pins. Which (in my mind) validates the "start over strategy" I am now taking; having recently ordered another new AMG8833 thermal camera.

  I am pleased to have attempted connecting the AMG8833 IR thermal sensor to a second computer. Because it helps me isolate where mistakes were made.

• 3 Months Asleep, QSP Has Been Reanimated

  Bret Bernhoft • 02/11/2026 at 01:36 • 0 comments

  After 3 months of hibernation the Quantifying Sleep Positions (QSP) project has been reanimated. In other words, my interest in self-quantification persists, so my various experiments do as well. Such as measuring how I sleep.

  In terms of hardware, I am restarting from scratch and taking a "blank slate" approach. For example, I have a new AMG8833 IR thermal camera from Adafruit on the way. I am also utilizing a recently purchased Raspberry Pi 5 computer.

  Where and how I failed in my previous attempt was by rushing through the setup process. Keeping such in mind, I am choosing to approach this project with greater respect and patience. Ultimately in search of premium outcomes.

  As progress is made, I will update this project with further log entries.

• Reconnecting AMG8833 IR Thermal Camera To Raspberry Pi 5

  Bret Bernhoft • 11/01/2025 at 09:06 • 0 comments

  As a matter of documenting my journeys in the world of Python and electronics, it seems prudent to update this project. In other words, I reaffirm my commitment to completing this endeavor as I have set out to do. In the meantime, I would like to provide a bit of context.

  I took a break from this project after hitting a roadblock and needing time to consider my next move. For example, in a previous project log I speculated about re-purchasing core hardware, fearing damage had been done to the original parts by me.

  With the above said and after some reflection, my next step is to reattempt connecting the AMG8833 IR thermal camera to the Raspberry Pi 5, in hopes of establishing communication between the two. My goal is to prove the assumption I mishandled electronics as incorrect. Once completed, I will post another project log here.

• Starting Over After Likely Damaging Hardware Components

  Bret Bernhoft • 09/13/2025 at 23:35 • 0 comments

  During my most recent attempt to assemble a RPi 5 and AMG8833 thermal camera, I was unable to make a working connection. In other words, the computer is unable to detect the AMG8833 sensor. Despite a dozen+ examinations of the wiring used and soldering done.

  While trying to get everything working, it seems I may have damaged the Raspberry Pi unit, as well as the IR sensor. As a result, the Raspberry Pi 5 does not boot. And the green activity light is a steady red.

  I am relatively unfamiliar with building electronics/devices, however simple they may be. I admit it, I am an amateur, which is good to recognize. It means I am curious and willing to put in the work to teach myself new skills.

  With all of that said, I am (most likely) going to repurchase both the Raspberry Pi computer and AMG8833 sensor. This is to help me avoid spending time on dead end rabbit holes, when I don't have the understanding(s) needed to debug this hardware yet.

  In some time I expect to look back on this project and see what my actual mistakes were/are. And laugh at this solution; to buy everything new and start over. Maybe it is for the best? For me to double down on this experiment and be willing to invest in its (in my) success. I am regardless.

• Beggining Anew And Making Progress As A Hardware Developer

  Bret Bernhoft • 08/29/2025 at 13:02 • 0 comments

  This project is my second Hackaday.io undertaking, for which I am glad to be in attendance. Efforts made here represent significant progress in my understanding of programming for "progressively exotic hardware and data visualization" scenarios, use cases.

  As an example, I had (again) set out to use a Raspberry Pi Zero 2 WH computer for this project. But my homelab wireless access point or WAP only accepts WPA3 connections; which the RPi Zero 2 WH cannot connect to/with. So I chose upgrading to a new Raspberry Pi 5; a computer capable of connecting to WPA3 WiFi signals.
  

  Ultimately I chose the RPi 5 model with 4 GB of RAM, being significantly more memory than requirements for the tasks this computer will be enabling.

  Current Hardware Situation

  As of August 29th in 2025, I have received all of the hardware components needed for building a prototype. And even attempted to assemble the monitoring device last weekend. But had limited initial success with connecting the AMG8833 sensor to the Raspberry Pi computer.

  Next steps include completing a functional first-generation monitoring device, capable of collecting and storing many hundreds of thousands of measurements. As well as beginning to work with (visualize using D3.js) said data generated while the contraption is operating over extended periods of time.

  As I make progress, I will update this project page.

View all 9 project logs