Death Stranding Odradek Scanner

Odradek Scanner

Oh no, there’s a BT in my study! The Odradek Scanner from Death Stranding, articulated version — you wouldn’t want to trigger a voidout, would you!

哔哩哔哩 ↗

Years ago when I played Death Stranding, I really wanted to replicate an Odradek Scanner. I searched the whole internet but couldn’t find a model I liked. Back then my modeling skills were terrible, so I had to give up.

Years later, while playing the sequel Death Stranding 2, I wanted to replicate the Odradek Scanner again. I opened the search engine once more, but the open-source community still had no model I liked — most were just non-articulated lighting props. Since an Odradek Scanner that doesn’t move is pointless, I started designing it myself.

Ch.0

The first thing to solve was the scanner’s articulated head. According to the art book description and in-game animation, the head is a rotatable structure, and the movable fins rotate in sync with the head. This would be very troublesome — it needs infinite rotation while driving multiple moving parts and lights, which is a challenge for both structural strength and electrical design.

Let’s solve the electrical problem first. If a part needs to rotate continuously while maintaining electrical connection, the best choice is a slip ring. I also used a slip ring in my previous micro nuclear power plant project. Slip rings come in many forms; the most common is this non-hollow type.

But the head needs to be driven to rotate. If I use this kind of slip ring, I’d need a hollow split BLDC motor to drive the head. Hollow split BLDC motors are very unfriendly for FDM 3D printed works (installation clearance / structural strength).

There is also a hollow conductive slip ring on the market (as shown below). The drive shaft passes through the middle, and other electrical signals are transmitted through the conductive slip ring, enabling infinite rotation while maintaining electrical connection.

With the electrical path solved, next is driving the rotating head (hereafter called the main shaft). How to drive it — use a servo or a BLDC gimbal motor (low KV)? The biggest advantage of a BLDC gimbal motor is silence; it makes almost no sound when rotating, while a servo produces noticeable noise. At the start of the design, I planned to use a BLDC motor to drive the main shaft. But after days and nights of searching and countless combinations of slip rings and motors, I never found a suitable BLDC motor for the main shaft — either the KV was too high or the size was wrong (the rear of the scanner has a flat profile). In the end, I compromised and used a 360-degree servo. (Custom motors are really expensive, I just can’t afford one QwQ)

Now the last problem: how to position a 360-degree servo. Unlike a normal servo, the duty cycle of the input signal controls the rotation speed, not the angle. But the head needs to find its zero point — every time it stops rotating, it must return to a fixed position to continue operating.

So I installed a reflective infrared sensor at the rear of the head and designed a bump at the head’s zero position. When the head returns to zero, it triggers the infrared sensor and notifies the MCU that homing is complete.

At this point the head solution is done. It seems the most troublesome part is solved.

Ch.2

For the movable fins I chose 9g servos — no reason, they’re cheap. Since the main shaft already uses a servo ┑(￣Д ￣)┍, a bit of noise is fine.

Since the conductive slip ring only has 4 electrical lines, only 2 can be used for signals. One is for LED control, one for the servos, meaning all servos must run on a single signal line. This made servo installation direction tricky — the two left fins must move opposite to the two right fins. The final design mounts the servos on top of the fins, installed in different orientations.

And Pidanlong designed a PCB for each fin — thanks to Pidanlong (https://space.bilibili.com/1933970) for the support

The LEDs use WS2812, single-wire RGB LED beads. One signal line controls all LEDs, and the lit effect is quite good

Ch.3

With the main parts solved, the rest is the overall appearance. Referring to the original art and design drawings in the art book, I completed the remaining exterior design

Assembled the body for a test — the carbon fiber panel print looks really nice

Ch.4

The main control SoC is still our familiar Espressif ESP32S3. The software implements three modes of the Odradek Scanner and uses a built-in web page to switch modes. I wrote code for servo control / built-in webpage / homing calibration, etc.

The Odradek Scanner has three modes in total

1. Scan mode: fins glow blue and slowly open/close periodically
2. Alert mode: fins glow orange and open/close rapidly
3. Alarm mode: fins glow red, top fins retract, surrounding fins move forward and rotate

Programmed with IDF5.3, code open-sourced on Github https://github.com/Hades2001/OdradekScanner

Ch.5

Of course, I can’t forget to write the assembly manual and usage (assembly is super complex, the 93-page assembly guide nearly killed me). The assembly guide is also open-sourced on Github https://github.com/Hades2001/OdradekScanner