Lucy's TODOList

This page is dedicated to a collection of all ideas being worked on for the Project Lucy fursuit. It is (hopefully) continually updated with a list of what we are doing in the project, but also serves as a pool of potential ideas or other ruminations~, for those interested in what is yet to be done.

Below you will find the ideas, sorted under "Active Construction" for ideas currently being developed, "Confirmed" for ideas and things definitely being added to the suit, and "Daydreams" for a low-urgency list of ideas that may or may not be added, and were simply fun to think about :>

Each sub-list is sorted further by the category it belongs to, such as hardware or software idea.

Changelog

A short summary of recently added/removed items to make it easier to keep up:

• 2024/12/20: Added idea for beamforming ear-microphone units.
  Added idea for gesture-based input/interaction.
• 2024/10/05: Added fake projector idea
• 2024/09/30: Redacted idea for custom protocol alternative to T1S. Pico PIO was not deemed capable enough to emulate the needed collision detections, and T1S MAC+PHY are cheap enough at ~5€/pcs.
  Added Front camera unit
  Added Wearable control panel.
• 2024/09/29: Added Microphone Output and DMX Input ideas; Telegram bot.

Active Construction

Hardware

• D.Er.G. Boards: These consist of the main LED driver and sensor controller boards, responsible for handling the LEDs, Sensors as well as a few audio effects:
  • Power regulation, providing ~3.5V to 5V for LEDs, Fans, etc.
  • Communication: DERG boards will use T1S Ethernet to network and transfer data to the main controller. This allows for up to 10MBit/s shared-medium data rates. On top of this will be a custom "lean" IPv6 Stack, probably reliant on pure UDP too.
  • LED Driving: DERG boards shall receive LED rendering data from the main controller and output it
  • Sensor handling: DERG boards shall include fault monitoring (Voltage/Current monitors), body monitors (Temperature, Humidity, more?), movement tracking (6 or 9-DOF IMU compatible with Slime-VR-like body movement tracking) and optional environmental sensors ( capacitive touch, LIDAR/Proxy, colour via TCS3472, microphone?)
  • Audio playback via a small driver and Speaker combo. Audio will be provided either via a stream from the core unit, or generated on the board itself.

Confirmed Concepts

Software

• Voice Recognition: In prior tests, CMU Sphinx // Pocketsphinx has proven very reliable as a fast, offline/standalone, Grammar-Restricted voice recognition. Grammar restriction here meaning that only a pre-defined list of utterances is recognized. This majorly improves precision and speed for audio input which only contains these preset utterances, and makes it viable for real-time voice control of the suit, handling emotions, etc.
• Telegram Bot: An interesting option to let people interact with the suit would be the use of a Telegram bot, which would do varying things such as receive updates about the suit's systems, upcoming events, or receive live location information to more easily find the suit. It could also contain interactive systems such as quizzes or similar. One fun idea was to let people send voice messages which are then spoken out of the suit's speakers!

Hardware

• Core Unit: The core unit will take care of the bulk processing and rendering tasks. It will provide the stream of Audio and LED data to be displayed/used, based on pre-made patterns and sensor data. The current target platform for this is a carrier board for a [LattePanda Mu](), which is a sufficiently powerful x86 SOM.
• Battery packs: The batteries for Lucy will consist of two or more Sodium-Ion based packs. Each pack will contain NaIon 18650 cells in series, with dedicated balancer and buck/boost charge/discharge controller. NaIon is chosen as technology thanks to its improved safety over Lithium-based cells, including safe and repeated discharge to 0V. To extend the operating range of the lower-capacity NaIon technology, at least two packs will be used, which shall be hot-swappable and shall alternate in discharging (such that one Pack will first discharge to 0% while the other remains at ~100%)
• Voice Modulator: A dedicated board for voice modulation shall be used to deliver near-real-time, smoothly modulated voice. This module should be as stand-alone as possible to allow other people to reuse it for their own purposes, and shall thusly include both a microphone and a speaker driver.
• Fursuit Head audio assist: To make communication easier on the suit wearer and listeners around them, we want to create a set of stand-alone hardware pieces that can be used to help. This shall consist of in-head headphones and, most importantly, beamforming microphones. These mics shall allow a much more directed communication with people directly in front of the suit, and thusly reduce influence from other directions.
• Ethernet Harness: To facilitate flexible and high-speed data transfer rates, a 100Base-TX ethernet network will be used. This network consists of a few small data and power switch-points, which handle measuring power draw for the branches and also translate the Ethernet connection to the 10Base-T1S of the DERG boards.

• Eyes: The eyes are probably one of the most important and expressive details of a Synth or Protogen suit. In the case of Lucy, the main eyes will be made of extra-bright LCDs running a fake ASCII terminal, into which the eyes are rendered, with the rendering being handled by a dedicated MCU (STM32H7 or ESP32P4). The eye rendering shall be procedural, so that it can directly react to eye tracking and other effects.

  In addition to LCD eyes, an idea that needs to be explored is hot-swappable eye modules. LCDs have highest level of detail but low brightness compared to RGB matrices, which would be better suited in brighter environments. For brightest environments such as in the sun, E-Ink or reflective LCDs shall be explored.

• Microphone output adapter: An adapter that interfaces directly to the suit's internal voice-changer output data, and can provide it via standard microphone or audio equipment connectors, so that it can be plugged into audio installations. It might also be of interest to use wireless standards for this. This would be awesome for Karaoke or similar events :D
• DMX Input unit: Similar to the mic output, except as input, for DMX data. The suit could have a few pre-set DMX-controllable values, so that it can be controlled/synced up to installations. See also the PawDMX workgroup on Telegram :D
• Wearable control panel: The chosen core unit, the LattePanda Mu, has the ability to connect directly to a small touch-panel. It may be useful to attach such a panel in an easily reached location on the suit, such as an arm, in a location that allows for status monitoring and configuring of the suit.

Suit construction

The suit construction of Lucy is still highly WIP, and involves various ideas for how to properly set up a comfortable, wearable suit - partial first, then Fullsuit.

For the initial partial suit design, the following pieces are conceptualised:

• A regular undersuit/glove/fabric liner, worn directly on the skin. Breathable, mainly for comfort to provide a soft surface.

• TPU Shaper: This element sits directly above the undersuit, and will take the role of "stuffing", to give the correct shape of the suit. It will be made of 3D printed TPU in an infill-only construction, giving a very flexible, very air-breathable material that can be printed to precise shapes and holds said shapes very well.

  The shaper will additionally contain air-channels, to guide air blown in by fans, LED channels for the under-fur LED lighting, as well as mounting locations for the shell pieces.

• A liner of fur above the TPU shaper, to provide a comfortable, huggable and soft outside. The fur will also act as light diffusor for the LED strips embedded into the shaper.

• Shell pieces: These shell pieces will either sit directly atop the internal shaper, attached via strong magnets or screws; or alternatively will use rubber bands or magnets to be worn above normal clothing in the case of a partial suit. These shell pieces will use a dual-material 3D printing process of clear and black high-hardness TPU, to create detailed external mechanical shapes of the Synth suit with embedded light-guiding patterns. The shell pieces also serve as attachment points for the DERG LED Driver boards, wiring, as well as the main portion of LEDs, an optional fan for air cooling, and a speaker.

• For the head: Strong air inlet/outlet fans to prevent fogging. The noise created by these fans should be mitigated by use of appropriate microphone+speaker combinations, both for outgoing and inbound audio. Audio handling will hopefully use a phased microphone array to make listening much more selective. With any luck we can even experiment with active noise cancelling?

  • The head might be constructed of two pieces: An internal head-support, which will include all the wearer-specific hardware (FPV, headphone, etc.), onto which the rest of the head couples?

Behavioral

"Behavioural" ideas are concepts that affect the aesthetic/visual/audio design of the suit, and are in essence parts of the act of the suit or design concepts to work alongside.

• Motor sounds: By using a combination of body tracking sensors (IMUs on the DERG boards) and audio playback (speakers on the DERG), it should be possible to create "fake" motor sounds. For this, the movement of a certain body joint (e.g. elbow) would be estimated, and according to this simulated sounds for different motor types could be generated, such as Stepper motors, Servos, Hydraulic/Pneumatic systems.
• Voice Modulation: One of the more important aspects of the suit will be a live voice modulator unit in the head, to provide real-time SFX. The exact specification of the voice modulation is not yet decided, but will most likely be made up out of a Vocoder stage, potentially phoneme recognition and usage of different samples in the Vocoder "harmonic to audio" conversion step. The end goal is to create a sufficiently smooth but artificial robotic voice (think GLaDOS).

Daydreams

Hardware

• Snoot sensors: A small assortment of sensors in the main head, for a variety of purposes. These could include a small camera module to take photos/videos (not for AR, just recording); a distance sensor to keep track of, and react to, boops, and also adjust the focal distance of the eyes to be appropriately silly when someone holds a hand close to the face.
• Front camera unit: A small camera unit seated in the visor of the suit head. This camera could be used for different purposes - taking pictures of people being one, but potentially also be used for effect handling, detection of wave gestures or suits :D
• Sensor projector unit: A small, cheap projector, either shoulder or face-mounted, to project out fake "scanning" lines :D

Effect ideas

• POV-Bar: A high density LED strip or bar attached to the arm or side of the suit, whose purpose it is to act like a POV display. The functionality would be such that with rapid waving of the arm, the strip could draw out patterns and words in the air around the suit.

Behavioral

• Proot-To-Proot: An idea for an audio-based contact detail exchange format, based on harmonic chords to encode data, mainly to be used as a system to let Synth and Protogen suits exchange details with one another or pretend to synchronize up with one another.

• Talk to the hand: An idea where a small speaker and microphone combination are placed on the hand, to turn it into a functional headset, allowing someone to, quite literally, talk to the hand.

• Gesture input/Silent Suiter mode: A mode where we do not use voice control to interface with the suit, but instead use the full body tracking of said suit to use hand-gestures as inputs. This could be akin to ASL, or perhaps coded to look like casting small spells//writing runes in the air? An example would be writing a little checkmark into the air and the suit playing a "Yes" message and animation, or other writings for "No", "Hello", "Hug?", etc

Scenes/Acts

• Cheese'd: A dedicated scene to be used when a piece of cheese is applied to the visor. Effects would include glitching out and repeating the word "Cheese" in more and more garbled audio output until the piece of cheese is removed.