Slack channel: #proj-breezy
Check figma to browse the ideas quickly.
This is a sheet with the relevant parts.
The project source code is public on github.
We have participated in the Code Life Ventilator Challenge.
The latest video
Problem and Background
During an outbreak of a virus causing respiratory problems, many patients must rely on mechanical ventilation. There may not be enough such devices in the peak of an epidemic. This would inevitably lead to loss of lives. There is a global need for a number of ventilators that cannot be supplied by commercial manufacturers in such a short time. In such a desperate situation, we are looking for a viable solution. Many teams over the world have started trying to design and build a prototype of a ventilator that could be used on patients with ARDS. The requirements on such devices are tough.
Several requirement documents have been published. A good example to follow is the British Medicines & Healthcare products Regulatory Agency specification for a Rapidly Manufactured Ventilator System calls itself a “specification of the minimally (and some preferred options) clinically acceptable ventilator to be used in UK hospitals during the current COVID-19 pandemic caused by SARS-CoV-2 virus.”
Solution summary in simple terms
Most of the proposed devices we have seen would not be helpful at ARDS treatment. We are not trying to bring a brand new concept at all costs, as we don’t believe there is a bigger chance such an approach could lead to anything viable. We want to create a state of art open source ventilator utilizing best practices. We take into account that the parts, materials and manufacturing methods used should be widely available.
Thanks to the solution being open source and modular, it is easy for any other ventilator project to reuse the code and design. A good example is our Arduino based measurement and statistical system, which needs only 2 sensors in order to calculate 12 parameters.
An Android (possible iOs) app extends the display showing on-line charts, thus greatly upgrading the diagnostic performance. This is open as well and may be reused by other teams.
Solution summary in technical terms
A DC blower fan will be used as an air pressure source. We rely on a pressurized oxygen source. Flow controller modules based on servo valves with feedback from flow sensors will be used or developed to deliver the required mixture of air / oxygen (FiO2) at preset pressure / volume. Expiration process including PEEP pressure will be controlled by expiration servo valve and flow sensor. These will be easily detachable and autoclavable for the purpose of sterilisation. Flow and pressure sensor measurements will be further processed in real time into statistical numeric form, presented on a LCD screen. Additional real-time graphic representation will be available through an Android app.
State of advancement of the project
We have purchased flow and pressure sensors and implemented them into the controller. Based on measured airway pressure and flow, we have a calculation of the common parameters like PEEP, Pmean, Pmax, respiratory rate, inspiration and expiration volume and time etc. We have an LCD showing these parameters on line. We have a working Android app showing live graphs of the quantities. We are waiting for delivery of a DC blower fan for testing and implementation. We have a concept of the inspiration valve, we will machine it using our CNC router soon.
- Week 1
- Continued development of firmware, software
- Order the needed material and parts
- Week 2
- Continued development of firmware, software
- Manufacture prototype valves
- Week 3
- Obtain parts (2 week lead time)
- Build the alpha prototype (hardware)
- Start coding & tuning control algorithms
- Week 4
- Verify the prototype
- Build the beta prototype
- Consult with medical expert on mechanical ventilation
- Week 5
- Publish an article
Please provide a projected project timeline in bullet point form.
(we will prioritize projects that can deliver in a short term period)
Air and oxygen sources, metering principle
We plan to use a blower fan as the air supply. The DC brushless blower fan is capable of enough flow and pressure. The pressure can be adjusted by altering the rate of rotation. Alternatively the flow of air will be mainly controlled by a custom designed proportional valve and flow meter as a feedback.
The ventilator will rely on pressurized oxygen gas supply. The supply pressure is between 3 to 5 bars. Solenoid valves will be used for dosing oxygen. One possible concept is an industrial ready-made flow controller. We need to assess the availability and pricing of such controllers for oxygen and capable of required maximum flow. Alternatively, we may develop a custom flow controller, consisting of a PID-controlled proportional solenoid with oxygen flow meter as a feedback. The third alternative is simply using two ordinary oxygen solenoids for dosing at higher frequency, then passing the gas through a dampening volume. This way we may skip the necessity for a o2 flow sensor, if we know the pressures and volumes. Another alternative is using only one valve for oxygen in pulse-width modulation mode. A flow impulse damping would be also necessary to filter the flow variations. Optionally an O2 concentration sensor could be used to verify or help control the mixture concentration.
The expiration will be controlled using a proportional valve similar to the one used to control inspiration. These valves are the subject of our development. The expiration valve and expiration flow sensor are parts coming into contact with the patient's breath, therefore they must be easily detachable and sterilizable. They will be mounted from outside of the ventilator.
FIG. Concept of expiration valve
The electronics - the solution is built on an Arduino Mega based open source 3D printer controller called RAMPS along with LCD board with encoder and buzzer called RepRapDiscount Full Graphic Smart Controller
We have decided to use it for these reasons:
- Widely available
- Low price
- Long life cycle
- Wide community support
These boards have many features we need:
- 3 power FETs (for controlling motor or solenoid valve)
- Stepper motor controllers (we may use stepper motors to control valves)
- 128 x 64 px LCD (fits many parameters we need to show)
- Rotary encoder (to control lower priority parameters buried in the menu)
- Buzzer (in case the device needs attendance)
- Many free pins to connect other devices
- Superior high-impedance ADC (can directly read appropriately scaled analog sensors, pots)
We don't suggest to build the actual medical device on Arduino prototyping board. It is only for development and prototyping stage of the project. The final product needs to adhere to relevant standards and quality.
We expect to build a prototype of a ventilator, that fulfills all the necessary (must have) criteria on such a device - see Rapidly Manufactured Ventilator System. We publish our results on a regular basis. Any stage of our progress may be helpful to the others.
Safety, quality assurance and regulation
Regarding safety requirements we plan to meet the requirements of the British Medicines & Healthcare products Regulatory Agency specification for a Rapidly Manufactured Ventilator System.
To protect from over-pressure the ventilator will have a mechanical over-pressure vent set at 40 cm H2O.
Alarms (sound and light) will be implemented to alert personnel of any uncommon error, leak, disconnection, power loss etc.
A possible list of tests was elaborated here:
We have a medical expert experienced in mechanical ventilation to consult the feasibility.
The device itself will need to have power-on self tests, which need to be performed before usage.
We don’t have any plans to manufacture more than prototype pieces. The outcome shall be documented knowledge and open source plans, software and firmware.
Impact, issues and risks
We want to provide a good and verified design of a possible lung ventilator that could be clinically acceptable. If successful, this may have greater impact in developing countries, as the price of making the device may be a fraction of a commercial fully-featured one. A proven open-source code and drawings may encourage new, even local entrepreneurs to enter the market. More competition leads to broader availability and better price, allowing more people to get the medical care they need. Local manufacturers support the local economy, enabling even higher living standard in the country.
We will consider the project a success, if (more bullets = better success):
- We inspire others with our ideas
- Other people start contributing
- Other project will reuse a part of, or the whole design
- We successfully finish a viable prototype of a lung ventilator
- We publish an article describing our achievements
- We inspire somebody to adopt the design, get certification and start manufacturing
Known issues, potential risks, grey-areas:
- Availability of critical parts, like oxygen valves, blower fan, flow sensors
- Ability to control the air and oxygen flow and pressure with required accuracy (we have not started working on these yet. We need to get parts first)
Adam Krovina - mechanical designer, firmware programmer, electrical safety, team leader
Bill Foote - Senior expert Android application developer
Maros Lukac - mechanical designer, 3d prototyping, procurement
Juraj Marak - mechanical engineer, Android application tester
Csaba Ruman - firmware programmer, control engineer
Csaba Boros - mechanical engineer, oxygen vs material fire safety
Jakub Sulgan - medical expert in mechanical ventilation
Slovak organisation for space activities
The project has been funded by some of the members of the team and some of their friends so far. We are using some material we already have.
For the project we have bought the following:
Along with shipping it was EUR 174.23 including VAT.
We need to buy the following:
How much we currently need is 711 EUR.
Some prices are estimated with fast shipping (7 - 15 days premium). USD/EUR exchange estimated. Excluding VAT.
If you wish to follow the links or copy text from table, you can find it in this document, tab "shopping list".
- Short Name: #Breezy
- Created on: March 21, 2020
- Last update: July 12, 2021
- Grant information: Received €711.00€ from the OpenCOVID19 Grant Round 1 on 04/07/2020