Introduction: SimpleClean Psycho Pass Dominator Prop

About: Things made with crappy tools, an impatient maker and in an imperfect manner in a rush, but just done well enough to call projects

This is my first attempt in prop making. I managed to rush this out in a week of school holiday, through building on the spot.

This dominator prop to be able to transform from a Non-Lethal Paralyser into the Lethal Eliminator Mode, play the appropriate sounds. It can also detect people's faces to give a crime coefficient, though not very fast. There are NeoPixel LEDs too!

This could be a good holiday project to not only improve prop-making skills but also have electronics & programming skills. It even has a camera which you can try some computer vision.

Step 1: Casing

The casing is made with a combination of MDF wood, Clear PVC plastic & Black Paper as covering.

First, the MDF wood was cut with a penknife, scissors, and a coping saw. The top MDF pieces were glued together.

The clear PVC plastic was cut with scissors and bent with pliers. There are 2 big sides to keep the general structure of the prop stable, and many small bent pieces of PVC to secure the 2 big sides together through hot glue.

Step 2: Transform Mechanism

The mechanism was made with 2 servos, one for the rectangular panels, and another for the cuboid (Housing the Raspberry Pi & other electronic components)

The servo for the rectangular panels was velcroed onto the cuboid. The servo moving the cuboid was secured to the cuboid by drilling holes into the plastic & the servo arm and connecting them with zinc galvanised wire.

Step 3: Securing Materials & Components

Black Paper was used to cover the exterior MDF wood pieces.

The electronic components and the plastic pieces in the internal frame were secured with hot glue & some velcro. The exterior wood pieces were glued together and secured to the internal frame with rubber bands & magnetic tape (Good for keeping the frames flushed with the frame.)

All of the electronic components were connected together through header pins. Some soldering was done, to wire up the Raspberry Pi's GPIO to the layout required by Servos (as a mini PiHat). The NeoPixel LEDs were wired up to 5V, GPIO 18, and GND. The Servos were wired up to 3.3V, GPIOs 17 & 27 and GND, while the roller switch (acting as a trigger) was wired to 3.3V and GPIO 24.

Step 4: Electronics & Software

The device has a Raspberry Pi Zero, connected to a camera, Roller Switch (to act as a trigger, on pin 24), NeoPixel LED strip (about 60 LEDs, on pin 18), 2 Micro Servos (pin 17 and 27). There is no in-built battery, with the whole setup powered by a micro-USB cable. The Raspberry Pi Zero connects wirelessly to a Bluetooth Speaker, to play different Psycho-Pass Sounds.

The Raspberry Pi is set up as a USB Ethernet Gadget, with SSH, VNC & Camera enabled, like in https://desertbot.io/blog/ssh-into-pi-zero-over-usb

The code running on the Raspberry Pi Zero is in Python and runs on boot-up by putting the commands in /etc/rc.local. It turns on the NeoPixel LEDs to cyan (like the colour of the Psycho-Pass Dominator Lights), and plays different Crime Coefficient Sounds on pressing the trigger. It uses these libraries:

  • gpiozero (Built-In)
  • rpi_ws281x adafruit-circuitpython-neopixel (you need to run "sudo pip3 install rpi_ws281x adafruit-circuitpython-neopixel")

The code used would be available soon (1 month's time). Same for the next page

Step 5: Computer Vision

The Raspberry Pi Zero has the potential to run face detection software, so that it could give Crime-Coefficient readings from a person's face.

Instead of compiling OpenCV, I download & installed a pre-compiled binary like in https://www.pyimagesearch.com/2018/09/26/install-opencv-4-on-your-raspberry-pi/. A good resource to start with Raspberry Pi Computer Vision is https://circuitdigest.com/microcontroller-projects/raspberry-pi-and-opencv-based-face-recognition-system.

Step 6: Conclusion

Overall, for a short holiday project, it is quite worthwhile to do. I made it in time for Anime Festival Asia 2019.