Loading 3D model… large boards can take a moment.
AKA the GPAD.
Development of the GPAD is in collaboration with the Sustainable Progress and Equality Collective (SPEC) and other individual volunteers.
GPAD in Prototype Enclosure of Summer 2022
The GPAD Printed Circuit Assembly Version 1 This General-Purpose- Alarm-Device repository defines a 'General Purpose Alarm Device' aka, GPAD module. This module has an enclosure and inside is an embedded microcontroller system with the following features to alert a human of a condition.
The GPAD does not detect the conditions that need to be alarmed. It is a peripheral to a controlling host device. The controlling host will assert status through the GPAD to an end user.
CAD view of GPAD With Version 1 Enclosure.
User view of GPAD Version 1 Printed Wiring Assembly without enclosure.
View of GPAD Version 1 Printed Wiring Assembly component side.
Files we used to build the V1 assemblies. PCB with some SMT Assembly was by JCLPCB in China. Additional parts were ordered and hand solderes in the USA. FILES HERE: Hardware/Manufacturing
The Freespireco Project is an attempt to build a modular ecosystem of cooperating respiration devices. One thing that devices such as ventilators and oxygen concentrators need to have is to be able to alert someone to dangerous conditions. This can be as simple and gentle as, "it is time for scheduled maintenance," to a far more dangerous, "the hose connected to the patient is disconnected!" The purpose of the alarm is to get the attention of nurses or doctors who can fix the problem.
See the PolyVent project on Gitlabs at: https://gitlab.com/polyvent
See the PolyVent controller at: https://gitlab.com/polyvent/polyvent_control_module
GPAD offers a simple API, possibly with a JSON or byte-level interface, for commanding it to alarm. Abstractly, we can imagine a system that supports five levels of alarm severity. Additionally, an alarm condition normally has a short message associated with it, such as "hose disconnect", which should potentially be displayed on a small screen.
The initial version of this API can be extraordinarily simple: Alarm at a level between zero and five, where zero means no alarm. However, it is clear this API will have to evolve over time. Issues such as when one alarm level supersedes another, when it is acceptable to silence an alarm and for how long, etc. actually become rather complicated quickly.
We hope that by defining a clearly versioned API we can make an evolving device of great utility for a wide variety of purposes.
Our fully finished design includes a printed wiring assembly and an enclosure. All the free and open source KiCad schematic printed circuit assembly files are provided so that users could fabricate just the PCB and use in an enclosure iof their own design. The enclosure is designed in the free open source SCAD and we provide the enclosure design source files. Users could fabricate this enclosure as is or develope a modifed enclosure with their own features from the files we provided.
The implementation is Arduino UNO like. A USB interface to an ATmega328 controller is provided for Arduino IDE compatibility. UNO footprint for shields is provided. This allows a typical Arduino serial port interface for development and debugging.
The wired RJ12(6P6C) Data connection from a host is an SPI controller and the GPAD is an SPI Peripheral.
A wireless interface may be added through a UNO shield. (Compatibility with existing wireless shields has not been checked. A custom shield with a wireless interface to some of the unused controller pins should be possible but may not be trivial.)
Demonstration of SPI controller and Peripheral for single byte transmission.
Still remaining: define a simple API. We imagine these hardware choices will evolve over time. We want the API to insulate the user of the GPAD from specific details of how the noise is made.
Nonetheless, we can make a few comments.
Summer and Fall of 2022 the physical implementation has been realized with the Version 1 PCB and Version 1 3D printable enclosure.
THESE MAY ALL HAVE BEEN REALIZED IN PHASE I
In Phase II, we can imagine a large number of improvements:
THESE MAY BE NESSISARY TO COMPLETE PHASE II
In Phase III, we could expand the usability of the device by allowing it to be made physically versatile, which feature such as:
In each of these phases, we imagine the API improving substantially.
Project Lead, Product Definition, Robert L. Read, founder of Public Invention.
Enclosure design, component procurement for rev1, project facilitator, initial board bring-up, board layout design contributor, Lawrence Kincheloe of Speccollective who is a computer scientist by training, and an electrical engineer as well as a hobby inventor by vocation.
??? Joe Torreggiani of Speccollective
Mentor and coach hardware, firmware, (Forrest) Lee Erickson Physicist who pretends to be electrical engineer, with experence in firmware development, Design for Manufactur, Design for Test and developer of Novel and Diverse Spelling.
We need volunteers:
Like everything done by Public Invention, this is a fully open project that will be released under fully open hardware and software licenses. However, unlike many of our projects which are very "researchy" this project could become a scalable product pretty easily. One can imagine it being sold at Sparkfun, Adafruit, or DigiKey if we do a good job. Because it could be used for a wide variety of purposes by makers and could be sold with an enclosure at a price of perhaps <=$100USD As an embedded sub assembly, the printed wiring assembly must be low cost with a material price of < $30USD
Potential uses as a product (by hobbyist) include:
Public Invention does not wish to become a manufacturer; but we will impartially support anyone who wants to take these designs and manufacture them so long as they abide by the licenses.
As this is intended to be a long-term project that will evolve and potentially have many volunteers. The basic software architecture is captured in this diagram:
The simplest API could just be:
enum AlarmLevel { silent, informational, problem, warning, critical, panic };
int Alarm(AlarmLevel al);
The simplest binding for this, which could be used of SPI or a serial port, would simply be the character A followed by the character 0, 1, 2, 3, 4, or 5. To test the alarm system of the serial port, you could simply type "A3" into the Arduino IDE serial monitor and send this to the GPAD, at which would expect a moderate noise and flashing light to sound and flash.
A slight improvement on this is to specify a message. We have decided that messages should have 80 character, 32 character, and 16 character versions, and the hardware device will do the best job rendering them in can with available hardware:
enum AlarmLevel { silent, informational, problem, warning, critical, panic };
int Alarm(AlarmLevel al);
int Alarm80(AlarmLevel al,char* msg);
int Alarm32(AlarmLevel al,char* msg);
int Alarm16(AlarmLevel al,char* msg);
Additionally, it would be nice to have a way to interrogate the current alarm level:
AlarmLevel getAlarm(AlarmLevel al);
boolean isMuted();
As of Nov. 11th, we are working significantly on this API, which deserves a a separate document, but I will place notes here.
The Abstract API is meant to work independent of specific hardware on the GPAD. That is, it should be useful for hardware built by other teams making significantly different decisions.
In this API, the message strings support only the most limited sort of formating: a linefeed character represents a line break. Nonetheless, if linefeed characters are used, the will break lines but this will of course limit the number of ACTUAL characters that can be rendered, rather than the 60 that we can render if no line breaks are used.
The Hardware API gives a caller complete access to all of the GPAD hardware. For example, they could address each LED separately. This is a lower priority. This API is not intended to be used by medical devices because it is fragile with respect to improvements in the hardware. Nonetheless, its availability makes the GPAD more useful for a wider variety of purposes. For example, someone could program a simple game with the GPAD.
In practice, there need to be three ways of invoking the API:
It is our intention to keep all of these mechanisms distinct from the two APIs. There might even be additional transport protocols in the future, such as a library that uses I2C ("I-squared C").
We have written in the "simulation" directory both the diagram and two sketches that simulate our hardware. One of them,
that Rob wrote on Saturday, August 13th, implements this rudimentary api, without supporting messages.
The other shows the use of the LCD.
These should be integrated together to create the initial firmware for the project. Meanwhile, we can simulate the device successfully in WokWi.
While the Version 1 PCBs and PCAs were on order, Lee wrote and sumulated factory test software on WokWi. See project at: https://wokwi.com/projects/345960281987351124
We're going to use white-light LEDs to avoid color-blindness issues for the visual alarm.
| Label Name | Short Description | Detailed Description |
|---|---|---|
| bug | Something isn't working | |
| documentation | Improvements or additions to documentation | |
| duplicate | This issue or pull request already exists | |
| enhancement | New feature or request | |
| firmware | For issues related to software running on the microcontroller | |
| good first | Good for newcomers | |
| Hardware | Schematic and PCB and Enclosure | |
| help wanted | Extra attention is needed | |
| question | Further information is requested | |
| research | Research products or ideas | |
| wontfix | This will not be worked on |
The firmware can be built with the standard Arduino IDE. (Using 1.8.19 as of 20221208) It requires libraries:
No comments yet. Be the first to ask about this board.