Building Your Own Randonaut Device

DIY Randonauting Device

Randonauting is an activity where random number generation is used as a tool for discovering and exploring nearby locations. The way it works is that random numbers are used to calculate the latitude and longitude coordinates of somewhere nearby and then you visit the real-world location.

There’s a metaphysical mind-matter aspect to this where your intentions are supposed to influence the randomly generated destination. The NY Times said it best – “Think: the Law of Attraction meets geocaching.”

Why would this work? Well, some believe that by using random numbers generated by quantum processes, e.g. a HWRNG device, it’s possible to mentally influence the chosen destinations. The result is the manifestation of some truly surprising, enlightening or even disturbing outcomes. Case in point, the alarming Tik Tok video where randonauting teens discover a suitcase with dead body parts.

Dujour Randonaut Device

This mind-over-matter premise might not be as far fetched as it sounds. There’s some surprising research that seems to demonstrate that it’s possible to mentally influence random numbers generated by quantum processes.

What happens if you focus on a specific intention precisely when the random coordinates are generated? That is what randonauting is all about.

There’s a few differences between my device and the other apps. First – I’ve added a feature: Time. In addition to calculating random geo coordinates I also calculate a random time for the trip. The idea being that it might be more meaningful to identify a point in both time AND space. Journey to a specific location at a specific time to maximize the experience!

The second difference is in the way that I determine the location. I use just two random numbers to calculate the geo coordinates. Other implementations include the concept of “voids” and attractors” which use statistical algorithms to determine the locations. Attractors are essentially a clustering of values that point to a geo coordinate while a void is the opposite (lack of points). In my opinion these techniques just introduce unneeded complexity.

Voids and Attractors

This is one of my more involved builds, so you’ll need to have some hardware and python expertise if you want to try this out. I call the device “Dujour” (in homage to The Matrix). To follow along you will need the following:

You might be wondering why use a hardware based RNG when a computer OS can natively create random numbers. Great question. Hardware random number generators use quantum physical processes to create truly random numbers while operating systems use an algorithm. Under the covers the OS based numbers are really pseudo-random. They’re random enough for most purposes, but numbers generated using a quantum process are truly unpredictable (at least in theory). Plus, if you buy into the underlying theory of Randonauting which involves mind-matter interaction, there’s that research seems to show that mental intention can only influence random numbers created by quantum processes.


Now you could build this device with a single Raspberry Pi by connecting the hardware RNG to a local USB port, but I prefer a separate device because I do a lot of experimenting with RNGs and it’s useful to have a remote RNG server that several devices can share.

The diagram below details the high-level Randonauting process flow. The primary script is called “” and is run on the first Pi, which I’ll call “Dujour1”. When you run it will make a REST call to the 2nd device “Dujour2” (the hardware RNG host), retrieve a few random float values, and then use those values to calculate the nearby location to explore along with the time to visit. The script then assembles a Google Maps URL and texts it to a phone via the Twilio service.

The video clip below shows the device in action.

Randonauts Device in Action

On the Dujour1 Pi, you’ll need to install and configure linux and connect your display. Follow these steps to connect a Matrix Orbital VK204-25. The image to the right shows the wiring for my display.

For my setup I housed both the Pi and the display in a bell jar and I connected a string of decorative LED lights to the 5V and ground pins on the Pi GPIO header. The jar was just a convenient way to hold it all together plus along with the LEDs I liked the aesthetic. 🙂

Dujour1 – Display Module Wiring

The script requires Python and the following libraries:

  • math
  • numpy
  • subprocess
  • sys
  • time
  • json
  • urllib2

If you get a dependency error when running the script you will need to install whatever module is missing.

There are several variables that need to be set prior to running. They’re all located in the script in the “User Defined Variables” section:

loghandle: path to a text file that logs all runs of the script

window_secs: Used to calculate the maximum seconds in the future to visit the location

meters_out: furthest distance possible for the geo coordinates in meters from your current location

latitude1, longitude1 = your current location (home base). This is used as the starting point

lcd_addr = hex address for LCD display if using I2C communications

HWRNG = IP address and port of remote HWRNG server. XXX.XXX.XXX.XXX:YYYY

There are a few dependencies on external scripts: is used to send the text message with the map coordinates. is used to drive the display. Place both scripts in the same directory as on Dujour1. Note: my script was developed to work with a specific Matrix Orbital display (VK204-25). If you decide to use a different one, you’ll need to change the code to work with yours. I’ve documented in the script where the interaction with the display takes place.

The script requires two OS environment variables to be able to authenticate with the Twilio service: ‘TWILIO_ACCOUNT_SID’ and ‘TWILIO_AUTH_TOKEN’. Follow these steps to configure the variables. You will also need to install the Twilio Python helper library.

On the “Dujour2” Pi you’ll need to install and configure Linux as well. This is where you will be connecting your hardware RNG. I used a OneRNG USB device, you can find the setup documentation here. (You can see my server in the image to the right.) Once configured, install and run the script to start serving up random numbers to Dujour1. Check here for detail on how the REST server script works.

Raspberry Pi Hardware RNG Server using OneRNG

If you’ve followed along up until this point, you should have everything you need to experiment with Randonauting using your own device. Just run ./ from a terminal and the result should be a text to your mobile phone with a map link (like the image to the right).

I’ve had some weird synchronicities when trying out my device. If nothing else, a random journey can open your eyes to nearby wonders that you’ve never noticed before.

In the future I might consider developing a custom Amazon Alexa skill. It would give me the ability to run my Randonauting server from my phone – wherever I might be.

Hey – drop me an email if you decide to build this. Let me know about your experience and any thoughts to improve the project!

Optimizing Psi with the Sidereal Pipe Light

Desk light that changes color based on sidereal time to notify you when you are most psychic

This idea is based on research that anomalous cognitive effects are more pronounced at specific sidereal times.

Sidereal Pipe Light

The premise of the journal article is that there are specific times of the day when you are more likely to be physic. This was based on analysis of experiments where participants attempted to “influence” random systems. By looking at the times associated with a database of 1468 free response trials, the author determined that anomalous cognitive effect size was 3-4x greater when occurring near 13.5 h local sidereal time.

Why use local sidereal time (LST) and not the standard time of the day?

Well, the author theorizes that there’s evidence for a casual connection between performance and the orientation of the subject or participant, the earth and the fixed stars.

What’s interesting is the characteristics of the effect size over the course of 24h LST.

It may be possible to increase effect size in AC experiments as much as four-fold by timing them near 13.5 h LST

On the other hand, you might not want to time your experiments 19h if you’re looking for a positive result.

Interesting to note is that Sagittarius A* (the supermassive black hole at the center of our galaxy) is right at the horizon at 13.5 h. One theory for the results is that Sagittarius A* is a significant source of background noise that dampens psychic effects and is most evident when directly overhead (around 19h LST).

Pipelight Build

For this project I programmed a Phillips Hue light to change color based on the expected effect size from the chart above. The idea being to move from one end of the color spectrum to the other (red to green) depending on whether the effect size was considered positive or negative.

Perhaps the most maddening part of the project was the complexity involved with developing an algorithm that could change the hue in even increments.

The chart above depicts the CIE colorspace which describes a coordinate for a given color.

Ultimately I just set the colors based on trial and error. I tested different color codes until I found an even transition for each hour in a 24 hour period.

I used python for this, I run the script on a schedule using crontab and send a command to the Hue bridge using the Phillips API.

Each hour has a code block that looks like the below. The “hue” parameter is the color address.

if LSThh == 22:
        if LSTmm < 30:
                puthue = '{"hue":12400}'
                print puthue
        if LSTmm > 29:
                puthue = '{"hue":14400}'
                print puthue

if LSThh == 23:
        if LSTmm < 30:
                puthue = '{"hue":15200}'
                print puthue
        if LSTmm > 29:
                puthue = '{"hue":14400}'
                print puthue

Another dilemma was how to determine LST based on my local time and location. Ultimately I borrowed this Sidereal calculator script.

I was very satisfied with the way the project turned out. Leave a comment and let me know what you think of this idea. If you’re interested in my final code, drop me an email and I’ll send it your way!