Embedded System for Monitoring Fuel Cell Power Supply System in Mobile Applications

Round 1

Reviewer 1 Report

Comments and Suggestions for Authors

Paper Ref : electronics-3564049

 

Review of :

“Embedded system for monitoring the FUEL CELL power supply system in mobile applications”

 

Author : Miroslav Matejček , Mikuláš Šostronek , Eva Popardovská , Vladimír Popardovský , Marián Babjak ,

 

Feedbacks :

Authors introduce a fuel cell (FC) based power supply system from Proton Exchange Membrane FC air/hydrogen reactants with an embedded controller monitoring system.  A 20 fuel cells stack is used in this study (ref FC H60) as possible mobile system applications. The interface is wireless communication (Bluetooth) for mobile platform or PC computer. The structure of this article is correct with design, curves and pictures of the organ system (quality of pictures could be improved). The references are well notified. At the end of the study, concrete experimental results are proven and could be an element for new exploration and development. This is an interesting study and please find hereafter some comments/remarks/questions to improve this first proposal:

 

 

To enhance this proposal:

• Please give all Unit(s) of equation elements in the text
• Remark : temperature is in Kelvin. Please update
• 1 and in the text lines [101, 102 … 115] precise the material type of electrodes (ide Anode & cathode)
• Fig2 and associated text [165 ….173] : please indicate the control pression/Tank of H2 and comment with the case of LH2. Which element controls the flow rate (Fig 3 and text)?
• Fig2 and for the picture : annotate all organs to be compliant with the schematic
• Clarify the INA N° in figure 3 ? (missing a label in the first one !  according to line 333)

In text and fig. 3 : what is the fuel cell controller (Arduino Mega ?) and comment T sens , purging llave , blower control .

• Please Annotate figure 5 (we see clearly the BLE module and …. ) to help the reader
• Why the load current starts @ 200mA (fig. 6 & 10) and starts @ 300mA (fig. 9) rather than 100mA with regular increase +100mA ? please comment in text.
• Remark : same in figure 12 (starts @ 200mA)
• For the transient FC behavior do you have evaluate the inductive case ? please argument in text .
• It seems that all measurements are done at room temperature (please indicate clearly in text); so what is the results at -40°C / + 65°C ?
• In the section 5) and in case of an update of the architecture , why do not use an STM32 µcontroller for example with all functions and embedded code? please comment
• Do the code is available as open source FC community?
• No added value for the Fig. 22
• In this system How is tracked a H leakage (safety condition)?

Comments on the Quality of English Language

ok for me 

Author Response

For Reviewer 1 (yellow higlighted parts in attached pfd)

Dear Mr./Ms./Mrs.

thank you very much for your review. We have try to answer/correct our article according to your reccomendations.

Please see text bellow and attached pdf file. In attached pdf file, the yellow higlighted text are our corrections according to your recommendations.

Please, the rest of highlighted text/figures are answers to others reviewers.

Thank you again for your time

 

To enhance this proposal:

• Please give all Unit(s) of equation elements in the text .... supplemented (please see attached file - yellow highlighted text)
• Remark: temperature is in Kelvin. Please update .... corrected in text
• 1 and in the text lines [101, 102 … 115] precise the material type of electrodes (ide Anode & cathode) .... supplemented in line 113...117
• Fig2 and associated text [165 ….173] : please indicate the control pression/Tank of H2 and comment with the case of LH2. Which element controls the flow rate (Fig 3 and text)? ... supplemented in line 198, and in Fig.2 line 201-205
• Fig2 and for the picture: annotate all organs to be compliant with the schematic... supplemented in Fig.2 line 201-208

• Clarify the INA N° in figure 3 ? (missing a label in the first one !  according to line 333) ... the information unused block has been deleted

In text and fig. 3: what is the fuel cell controller (Arduino Mega ?) and comment T sens , purging valve , blower control .

Figure 3. and text was corrected FC Controller (HFCT)

• Please Annotate figure 5 (we see clearly the BLE module and …. ) to help the reader

Figure 5 was corrected

• Why the load current starts @ 200mA (fig. 6 & 10) and starts @ 300mA (fig. 9) rather than 100mA with regular increase +100mA ? please comment in text.

Following text in result section was added:

 Measurement process starts with measurement at I_LOAD = 200mA, it continued at 100, 300 ... 700 mA. The main reason for this procedure was to restore and ensure stable operation of the PEMFC after a longer period of operational inactivity. The timeline in Figures 6-10,12 and their colour legend - description of the measurements demonstrate the above-mentioned measurement process.

• Remark : same in figure 12 (starts @ 200mA) ... Legend in Figure 2  was corrected
• For the transient FC behaviour do you have evaluate the inductive case ? please argument in text .

Text in the conclusion was supplemented: line 748-751:

The experiments did not assess the effects of connecting inductive, capacitive or mixed loads, which in UGVs are represented by the electric motors used or the on-board battery. The effect of the load also represents an area for further research.

• It seems that all measurements are done at room temperature (please indicate clearly in text); so what is the results at -40°C / + 65°C ?

Text in results section was supplemented:

All measurements were realized at ambient temperature 298.15 K.

Text in conclusion was supplemented:

PEMFC can also be tested using the VT7004 temperature chamber located at our department, which with its temperature range from 223 to 423 K covers the entire operating range of the PEMFC H60.

• In the section 5) and in case of an update of the architecture, why do not use an STM32 µcontroller for example with all functions and embedded code? please comment

Text in discussion section was supplemented:

 

The architecture of the monitoring system with the Arduino MEGA platform can also be created using 8-bit or 32-bit STM32 controller platforms, where the manufacturer (ST) offers a set of microprocessors for solving various application tasks. These are also intended for general use, for signal processing, for the use of machine learning ML and security applications.

The practical issue of creating new architecture of the monitoring system is associated with other questions such as: what development environment the platform manufacturer provides for developers, what number and types of interfaces does the development platform have available, what is the connectivity of the platform and wireless technologies and what is the compatibility of the platform and expansion modules from other manufacturers, and others. In case we would like to take advantage of the Arduino platforms with which we have broader experience, the simplicity of designs and the compatibility of existing modules, it is possible to lean towards 32-bit variants, for example, Arduino UNO R4 or Arduino DUE (with several UARTs).

• Do the code is available as open source FC community?

Text in discussion section was supplemented:

Proposed script used in the Arduino MEGA platform is not available as open source code in FC community, yet.

• No added value for the Fig. 22 … Figure 22 was deleted
• In this system How is tracked a H leakage (safety condition)?

Text in conclusion was supplemented:

During these tests, it will be necessary to monitor the concentration of unconsumed hydrogen in the temperature chamber using gas measurement modules (for example, the MQ-2, MQ-7 series or others). Many similar modules are compatible with the Arduino MEGA platform.

 The manufacturer's PEMFC controller (HFCT) and monitoring system in our article were not equipped with a hydrogen concentration sensor. Safety during operation was ensured by sufficient ventilation.

Author Response File: Author Response.pdf

Reviewer 2 Report

Comments and Suggestions for Authors

See the attached file. 

Comments for author File: Comments.pdf

Comments on the Quality of English Language

no

Author Response

To Reviewer 2

Dear Mr./Ms./Mrs.

thank you very much for your review. We have to try to answer/correct our article according to your reccomendations.

Please see text bellow and attached pdf file. In attached pdf file, the green higlighted text are our corrections according to your recommendations.

Please, the rest of highlighted text/figures are answers to others reviewers.

Thank you again for your time

M.

 

This article focuses on the research of embedded monitoring system for fuel cell power supply system in mobile applications. The description is clear, the experimental design is also reasonable. But some contents need to be improved.

1) The background of fuel cell technology is relatively comprehensive, but recent research progress needs to be supplemented (Note：In the introduction part, the related literature cited are Res.1-10 with the latest literature being 2019).

References 8-14, 30-31 were added and state of art was updated:

Introduction section was supplemented:

Analyse the state of art is based on research results mentioned by authors in [1-7]. Re-searchers at the present time focus on the discovery of prospective materials and manu-facturing methods to improve fuel cell performance and simplify components of fuel cells. Designs of fuel cell systems are utilized to reduce the costs of the membrane and improve cell efficiency, reliability and durability. Those improvements allowing them to compete with the combustion engine. Recent advancements as in [1] describe standard design off a grid-connected fuel cell system, advantages and disadvantages of different groups of fuel cells. Research in [2] is focused on improvement of electrodes, electrolytes materials of different types of fuel cells (polymer electrolyte membrane fuel cells, direct methanol fuel cells and solid oxide fuel cells). Authors in [3] describes a science and technology of a new type of electrocatalysts consisting of a single atomic layer of platinum on suitable sup-ports. This research can solve problems with price cost, activity, and stability—for a broad range of fuel cell applications. Authors in [4] deals with catalysts based on car-bon-supported platinum nanoparticles (Pt/C) for PEMFC. Main goal is analysed of meso-porous carbon structures and their influence on the durability of PEMFC catalysts.

Currently, the usage of embedded systems for monitoring various types of FC has been addressed, for example, in [5, 6, 7], where the author [5] deals with design of embed-ded monitoring system powered by FC, which is capable to measure/logg only FC voltage and monitoring system code is optimized to reduce its power consumption. Next authors as in [6] are focused on application of active Fault Tolerant Control (FTC) strategy and experiments with PEMFC. Author’s goal is to solve a (PEMFC) problem with water management/drying issue by coupling an Fault Detection and Isolation (FDI) algorithm, a re-configuration mechanism, and an adjusting embedded system (controller). The process FDI uses a neural network; then, a self-tuning PID is used as the control element. Next au-thors as in [7] used Electrochemical Impedance Spectroscopy method and its implementation in embedded system for monitoring of PEMFC fuel cells state.

 2) Some citation formats are inconsistent (such as missing page numbering in reference), and some reference formats are not standardized. Part of the cited content is redundant (such as multiple references to the same literature), it is recommended to integrate them.

According to recommendation of MDPI, we used Zotero software tool for references generation. Please, in case of yours next detailed recommendations, we will be able to correct references.

3) The calibration process of INA226 requires additional step explanations, and the communication protocol between the blue-tooth module and Arduino needs to be detailed explained.

Section 4. Results was updated:

The INA226 calibration process was started with the replacement of shunts accord-ing to the application requirements for measuring currents up to 20 A. The calibration it-self consisted of experimentally setting the resistance value of individual shunts (to 0.004 Ω) in the Arduino MEGA script. During the calibration, the current values between indi-vidual INA226 were monitored and compared. This methodology was applied for current values up to 100mA and up to 1A. The error in the current values measured by individual INA226 did not exceed 2.55%.

In Section Materials and Methodes was updated line 359-363:

Arduino MEGA has multiple UARTs for asynchronous communication. One of them was used for communication with Arduino IDE serial monitor and the second was used for communication with Bluetooth module HC-05. Data from INA226 modules was sent by Arduino MEGA via UART(RX/TX) channel to Bluetooth module HC-05. UART protocol and data-stream details are mentioned [18]. 

4) To enhance read ability, it is recommended to use flowcharts or sub-module diagrams instead of pure text descriptions.

Fig. 22 - Flowchart was created ..., please see in lines 586-587

5) In Conclusion, suggest give the shortcomings of the current research method, and propose future improvement directions

Follow texts were added in discussion section in line 611-664:

The architecture of the monitoring system with the Arduino MEGA platform can also be created using 8-bit or 32-bit STM32 controller platforms, where the manufacturer (ST) offers a set of microprocessors for solving various application tasks. These are also in-tended for general use, for signal processing, for the use of machine learning ML and se-curity applications.

The practical issue of creating new architecture of the monitoring system is associat-ed with other questions such as: what development environment the platform manufac-turer provides for developers, what number and types of interfaces does the development platform have available, what is the connectivity of the platform and wireless technologies and what is the compatibility of the platform and expansion modules from other manu-facturers, and others. In case we would like to take advantage of the Arduino platforms with which we have broader experience, the simplicity of designs and the compatibility of existing modules, it is possible to lean towards 32-bit variants, for example, Arduino UNO R4 or Arduino DUE (with several UARTs).

The proposed script used in the Arduino MEGA platform is not available as open source code in FC community. Final data analysis (efficiencies computing etc.) was per-formed in the MATLAB program environment.

The proposed monitoring system is expandable for large fuel cell stacks tanks to a large number I2C connected power monitors, thanks to high connectivity options of Ar-duino MEGA platform. 54 Input and output I/O digital pins and 16 analog pins could be used for the temperature, vibration, concentration of leaked hydrogen sensing functions. The proposed PEMFC monitoring system can also be used in other renewable sources of electricity generation, for example in hydro or wind turbines. The advantages of embed-ded systems include their functional reconfiguration, and the composition of the moni-toring system based on multiple external interrupt-controlled embedded platforms.

 Primary limitation of the overall PEMFC power supply system is the necessity of an external power supply capable of supplying a voltage of 13V and a current of more than 0.6A for the needs of the PEMFC controller FC H60. An interesting finding of the experi-ments is that the FC controller's consumption (IEXT) decreases with increasing value of the generated current (ILOAD) flowing through the load. Therefore, it can be expected that in the case of generating the maximum nominal current ICELL ≈IOUT ≈ILOAD ≈5A, the ratio between the current produced by the PEMFC ICELL and the current IEXT consumed by the PEMFC controller will be significantly higher.

            Primary limitation of proposed PEMFC monitoring system is given by for example by choosing a suitable embedded system Atmega2560.  From this point of view, the mon-itoring system is limited by the number of UARTs interfaces of the Atmega 2560 platform (4 UARTs) which can be used for Bluetooth communication. Next its (Atmega 2560) limi-tation is the number of I2C interfaces to which a finite number (128) of I2C - connectable devices, in our case INA226 power monitors. When the number of used I2C devices (INA226) the frequency of monitoring functions will decrease rapidly.  The Arduino MEGA platform is usable over a wide operating temperature range from 233K to 358 K.

Next limitation is represented by chosen power monitors type INA226 and its basic characteristics as was previously mentioned (1 LSB resolution is for shunt voltage 2.5μV, bus voltage 1.25mV). INA226 is equipped with 16-bits native ADC with minimal (one) conversion time 150 microseconds. When using the averaging options for consecutive measurements with INA226, the conversion time increases to up to 9 milliseconds. INA226 is capable of sensing PEMFC voltages up to 36 V (in hight-side/low-side sensing configuration). The INA226 is usable over a wide operating temperature range from 233K to 398K with current consumption down to 0.5 milliamperes.

Bluetooth module HC-05 limitation is given by its basic characteristics: Bluetooth version V2.0+EDR, 3 Mbps, 2.4 GHz, transmit power is +4dBm default bau-drate/configuration 9600/38400 8N1, range up to 100 m, receiver sensitivity -80dBm, communication voltage level 3.3V. During the experiments, it was found that the commu-nication between the HC-05 and the Xiaomi Redmi 12 smartphone was more than 100 m (in the building environment and direct radio sight), however, the maximum distance of wireless communication was not determined in our experiments. The HC-05 module re-quires an operating current of at least 30 mA.

The estimated costs of the proposed system is approximately 80€, but we expect that future tests associated with the integration of the monitoring system will also require ad-ditional costs.

Compared to embedded systems and their software development environments, we can state that the costs of Arduino platform solutions are several times cheaper than solu-tions using embedded systems from other manufacturers (STMicroelectronics, Analog Devices or NXP Semiconductor). Of course, a cost-effective solution also brings the neces-sity of compromises regarding selected technical parameters of cheaper variants of em-bedded systems development platforms.

It is difficult to compare the price of our designed monitoring system with its own functionalities for small UGVs with other monitoring systems that could have different features. Nowadays, we have not found any identical or similar monitoring system for comparison with our proposed monitoring system.

In conclusion, following text was added (line 716-744):

Current FC investigation methods provide a picture of efficiency and monitoring from the perspective of electrochemical processes in relation to new FC technologies. In our article, we offer a more detailed picture of the electrical efficiency of a complex FC sys-tem. This view is not often available for end users.

In our further research, we therefore focused on the application of embedded moni-toring systems for a complex FC power supply system.

Environmental impact of the proposed system can be assessed in terms of the energy required to produce hydrogen using the Hydrofill Pro station from the manufacturer PEMFC H60. This filling station requires a power supply of 10-19V and a power of up to 23W. The filling speed is approximately 3 liters of hydrogen per hour with a consumption of deionized distilled water of 20ml/hour. Hydrogen purity was 99.99%.

During the 23 minutes experiments 19.3 liters of hydrogen were used (not counting the hydrogen used for the initial restart of the cell). At a hydrogen tank filling rate of 3 l/hour, the required filling time would be 6.4 hours, and the energy consumed for the fill-ing equipment would be approximately 148 W·h. Another 8W or more was required from an external power source for the needs of the manufacturer's H-60 PEMFC controller. This value corresponds to 3.1 W·h.

The power consumption of the Arduino MEGA platform depends on the selected supply voltage, the number and load status of the I/O pins but does not exceed 0.5A (built-in protection function). At a supply voltage of 12 volts for the Atmega 2560, a current of half an ampere, the energy expended will be equal to 6.4 Wh.

The total energy consumption during our experiment was approx.. 162.4 W·h, of which 91% of the energy was consumed by hydrogen production. Each Watt per second represents one Joule.

If the IPCC-designated carbon emission factor is used to calculate the equivalent car-bon emissions as in [29], the CO2 emissions from water electrolysis-based hydrogen pro-duction powered by grid power, wind energy, and solar energy are 39.74 kg, 2.04 kg, and 0.68 kg CO2-eq respectively. Because 1kg hydrogen is 14,13 litre, we used 1.37 kg for our experiments which corresponds 54.3 kg CO2 for the filling system supplied from public electric-grid.

Comparison of proposed system with traditional power supply systems can be found in our previous research [30] when different voltage regulators were evaluated from the point of view of their efficiencies. There one can found that traditional continuous (LDO) regulator efficiency is in interval 30-50% but switching regulators efficiency reach up to 94-98%.  

Author Response File: Author Response.pdf

Reviewer 3 Report

Comments and Suggestions for Authors

The paper presents an interesting study on the development of an embedded system for monitoring a fuel cell-based power supply system in mobile applications, particularly for Unmanned Ground Vehicles (UGVs). The topic is timely given the growing interest in clean and sustainable energy solutions, especially in the context of reducing environmental impact and improving energy efficiency with the integration of Bluetooth technology for data transfer and the use of power monitors for efficiency.

Comments: 

- The comparison with the state of the art is not fairly covered. The paper would benefit from a more detailed discussion of how the proposed system compares to existing solutions in terms of efficiency, scalability, and applicability in real-world scenarios.

- The references are quite limited and need enrichment. The paper would benefit from a broader discussion of related work, including recent advancements in fuel cell technology, embedded monitoring systems, and mobile applications. 

- The discussion is limited. While the results are presented, there is insufficient analysis of their implications, particularly in terms of the system's limitations, potential improvements, and broader applicability. A more in-depth discussion of the practical challenges and future research directions would strengthen the paper.

- The conclusion is too long and somewhat repetitive. It could be more concise, focusing on the key findings and their significance, while avoiding redundancy with the results and discussion sections.

Questions :

- What are the primary limitations of the proposed monitoring system, particularly in terms of its accuracy and reliability in dynamic mobile environments?

- How scalable is the proposed system for larger fuel cell stacks or different types of fuel cells (e.g., Solid Oxide Fuel Cells)? Could the same methodology be applied to other energy systems?

- What are the potential future enhancements to the system, such as integrating additional sensors or improving the data processing algorithms? How would these improvements impact the system's performance?

- Given the focus on clean energy, how does the environmental impact of the proposed system compare to traditional power supply systems, particularly in terms of carbon footprint and resource consumption?

- What is the estimated cost of implementing the proposed monitoring system in a commercial UGV or other mobile applications? How does this cost compare to existing monitoring solutions?

- Has the system been tested in real-world conditions, such as in a moving UGV or in varying environmental conditions (e.g., temperature, humidity)? If not, what are the expected challenges in such scenarios?

- How easily can the proposed monitoring system be integrated with other onboard systems in a UGV, such as battery management systems or navigation systems? Are there any compatibility issues that need to be addressed?

While the paper is interesting , the comments and questions raised must be carefully addressed before the paper can be considered for acceptance. 

Author Response

Dear Mr./Ms./Mrs.

thank you very much for your review. We have try to answer/correct our article according to your reccomendations.

Please see text bellow and attached pdf file. In attached pdf file, the turquoise higlighted text are our corrections according to your recommendations.

Please, the rest of highlighted text/figures are answers to others reviewers.

Thank you again for your time

M.

Comments: 

- The comparison with the state of the art is not fairly covered. The paper would benefit from a more detailed discussion of how the proposed system compares to existing solutions in terms of efficiency, scalability, and applicability in real-world scenarios.

- The references are quite limited and need enrichment. The paper would benefit from a broader discussion of related work, including recent advancements in fuel cell technology, embedded monitoring systems, and mobile applications. 

The answers to these 2 comments are supplemented in Introduction and Literature Review section (please see colour highlighted text in pdf file).

Analyse the state of art is based on research results mentioned by authors in [1-7]. Researchers at the present time focus on the discovery of prospective materials and manufacturing methods to improve fuel cell performance and simplify components of fuel cells. Designs of fuel cell systems are utilized to reduce the costs of the membrane and improve cell efficiency, reliability and durability. Those improvements allowing them to compete with the combustion engine. Recent advancements as in [1]  describe standard design off a grid-connected fuel cell system, advantages and disadvantages of different groups of fuel cells. Research in [2] is focused on improvement of electrodes, electrolytes materials of different types of fuel cells (polymer electrolyte membrane fuel cells, direct methanol fuel cells and solid oxide fuel cells). Authors in [3] describes a science and technology of a new type of electrocatalysts consisting of a single atomic layer of platinum on suitable supports. This research can solve problems with price cost, activity, and stability—for a broad range of fuel cell applications. Authors in [4] deals with catalysts based on carbon-supported platinum nanoparticles (Pt/C) for PEMFC. Main goal is analyse of mesoporous carbon structures and their influence on the durability of PEMFC catalysts.

Currently, the usage of embedded systems for monitoring various types of FC has been addressed, for example, in [5, 6, 7], where the author [5] deals with design of embedded monitoring system powered by FC, which is capable to measure/logg only FC voltage and monitoring system code is optimized to reduce its power consumption. Next authors as in [6] are focused on application of active Fault Tolerant Control (FTC) strategy and experiments with PEMFC. Author’s goal is to solve a (PEMFC) problem with water management/drying issue by coupling an Fault Detection and Isolation (FDI) algorithm, a reconfiguration mechanism, and an adjusting embedded system (controller). The process FDI uses a neural network; then, a self-tuning PID is used as the control element. Next authors as in [7] used Electrochemical Impedance Spectroscopy method and its implementation in embedded system for monitoring of PEMFC fuel cells state.

The references were supplemented by 7 references.

- The discussion is limited. While the results are presented, there is insufficient analysis of their implications, particularly in terms of the system's limitations, potential improvements, and broader applicability. A more in-depth discussion of the practical challenges and future research directions would strengthen the paper.

The above-mentioned comment is identical to the reviewer's question number 1. Therefore, I will take the liberty of answering within the framework of this question. The answers to this question are supplemented in Discussion section (colour highlighted).

- The conclusion is too long and somewhat repetitive. It could be more concise, focusing on the key findings and their significance, while avoiding redundancy with the results and discussion sections.

Conclusion section was updated according to yours and other reviewer’s recommendations.

Conclusion section was updated according to your recommendations.

Questions:

1. What are the primary limitations of the proposed monitoring system, particularly in terms of its accuracy and reliability in dynamic mobile environments?

Primary limitation of the overall PEMFC power supply system is the necessity of an external power supply capable of supplying a voltage of 13V and a current of more than 0.6A for the needs of the PEMFC controller FC H60. An interesting finding of the experiments is that the FC controller's consumption (I_EXT) decreases with increasing value of the generated current (I_LOAD) flowing through the load. Therefore, it can be expected that in the case of generating the maximum nominal current I_CELL ≈I_OUT ≈I_LOAD ≈5A, the ratio between the current produced by the PEMFC I_CELL and the current I_EXT consumed by the PEMFC controller will be significantly higher.

            Primary limitation of proposed PEMFC monitoring system is given by for example by choosing a suitable embedded system Atmega2560.  From this point of view, the monitoring system is limited by the number of UARTs interfaces of the Atmega 2560 platform (4 UARTs) which can be used for Bluetooth communication. Next its (Atmega 2560) limitation is the number of I2C interfaces to which a finite number (128) of I2C - connectable devices, in our case INA226 power monitors. When the number of used I2C devices (INA226) the frequency of monitoring functions will decrease rapidly.  The Aruino MEGA platform is usable over a wide operating temperature range from 233K to 358 K.

Next limitation is represented by chosen power monitors type INA226 and its basic characteristics as was previously mentioned (1 LSB resolution is for shunt voltage 2.5μV, bus voltage 1.25mV). INA226 is equipped with 16-bits native ADC with minimal (one) conversion time 150 microseconds. When using the averaging options for consecutive measurements with INA226, the conversion time increases to up to 9 milliseconds. INA226 is capable to sense PEMFC voltages up to 36 V (in hight-side/low-side sensing configuration). The INA226 is usable over a wide operating temperature range from 233K to 398K with current consumption down to 0.5 milliamperes.

Bluetooth module HC-05 limitation is given by its basic characteristics: Bluetooth version V2.0+EDR, 3 Mbps, 2.4 GHz, transmit power is +4dBm default baudrate/configuration 9600/38400 8N1, range up to 100 m, receiver sensitivity -80dBm, communication voltage level 3.3V. During the experiments, it was found that the communication between the HC-05 and the Xiaomi Redmi 12 smartphone was more than 100 m (in the building environment and direct radio sight), however, the maximum distance of wireless communication was not exactly determined in our experiments. The HC-05 module requires an operating current of at least 30 mA.

The answers to this question are supplemented in Discussion section (colour highlighted).

2. How scalable is the proposed system for larger fuel cell stacks or different types of fuel cells (e.g., Solid Oxide Fuel Cells)? Could the same methodology be applied to other energy systems?

Proposed monitoring system is expandable for large fuel cell stacks tanks to a large number I2C connected power monitors, thanks to high connectivity options of Arduino MEGA platform. 54 Input and output I/O digital pins and 16 analog pins could be used for the temperature, vibration, concentration of leaked hydrogen sensing functions. The proposed PEMFC monitoring system can also be used in other renewable sources of electricity generation, for example in hydro or wind turbines. The advantages of embedded systems include their functional reconfiguration, and the composition of the monitoring system based on multiple external interrupt-controlled embedded platforms.

The answers to this question are supplemented in Discussion section (colour highlighted).

3. What are the potential future enhancements to the system, such as integrating additional sensors or improving the data processing algorithms? How would these improvements impact the system's perform

The answer to this question is given in the conclusion. (colour highlighted).

PEMFC can also be tested using the VT7004 temperature chamber located at our department, which with its temperature range from 223 to 423 K covers the entire operating range of the PEMFC H60. During these tests, it will be necessary to monitor the concentration of unconsumed hydrogen in the temperature chamber using gas measurement modules (for example, the MQ-2, MQ-7 series or others). Many similar modules are compatible with the Arduino MEGA platform.

 The manufacturer's PEMFC controller (HFCT) and monitoring system in our article were not equipped with a hydrogen concentration sensor. Safety during operation was ensured by sufficient ventilation at measurement site (laboratory).

 

4. Given the focus on clean energy, how does the environmental impact of the proposed system compare to traditional power supply systems, particularly in terms of carbon footprint and resource consumption?

The answers to this question are supplemented in Discussion section (colour highlighted).

Environmental impact of the proposed system can be assessed in terms of the energy required to produce hydrogen using the Hydrofill Pro station from the manufacturer PEMFC H60. This filling station requires a power supply of 10-19V and a power of up to 23W. The filling speed is approximately 3 liters of hydrogen per hour with a consumption of deionized distilled water of 20ml/hour. Hydrogen purity was 99.99%.

During the 23 minutes experiments 19.3 liters of hydrogen were used (not counting the hydrogen used for the initial restart of the cell). At a hydrogen tank filling rate of 3 l/hour, the required filling time would be 6.4 hours, and the energy consumed for the filling equipment would be approximately 148 W·h. Another 8W or more was required from an external power source for the needs of the manufacturer's H-60 ​​PEMFC controller. This value corresponds to 3.1 W·h.

The power consumption of the Arduino MEGA platform depends on the selected supply voltage, the number and load status of the I/O pins but does not exceed 0.5A (built-in protection function). At a supply voltage of 12 volts for the Atmega 2560, a current of half an ampere, the energy expended will be equal to 6.4 Wh.

The total energy consumption during our experiment was approx.. 162.4 W·h, of which 91% of the energy was consumed by hydrogen production. Each Watt per second represents one Joule.

If the IPCC-designated carbon emission factor is used to calculate the equivalent carbon emissions as in [29], the CO2 emissions from water electrolysis-based hydrogen production powered by grid power, wind energy, and solar energy are 39.74 kg, 2.04 kg, and 0.68 kg CO2-eq respectively. Because 1kg hydrogen is 14,13 litre, we used 1.37 kg for our experiments which corresponds 54.3 kg CO2 for the filling system supplied from public electric-grid.

Comparison of proposed system with traditional power supply systems can be found in our previous research [30] when different voltage regulators were evaluated from the point of view of their efficiencies. There one can found that traditional continuous (LDO) regulator efficiency is in interval 30-50% but switching regulators efficiency reach up to 94-98%.  

 

5. What is the estimated cost of implementing the proposed monitoring system in a commercial UGV or other mobile applications? How does this cost compare to existing monitoring solutions?

The estimated costs of proposed system is approximately 80€, but we expect that future tests associated with the integration of the monitoring system will also require additional costs.

Compared to embedded systems and their software development environments, we can state that the costs of Arduino platform solutions are several times cheaper than solutions using embedded systems from other manufacturers (STMicroelectronics, Analog Devices or NXP Semiconductor). Of course, a cost-effective solution also brings the necessity of compromises regarding selected technical parameters of cheaper variants of embedded systems development platforms.

It is difficult to compare the price of our designed monitoring system with its own functionalities for small UGVs with other monitoring systems that could have different features. Nowadays, we have not found any identical or similar monitoring system for comparison with our proposed monitoring system.

The answers to this question are supplemented in Discussion section (colour highlighted).

6. Has the system been tested in real-world conditions, such as in a moving UGV or in varying environmental conditions (e.g., temperature, humidity)? If not, what are the expected challenges in such scenarios?

The answers to this question are supplemented in conclusion (colour highlighted). The same requirement had Reviewer 1 (yellow highlighted).

7. How easily can the proposed monitoring system be integrated with other onboard systems in a UGV, such as battery management systems or navigation systems? Are there any compatibility issues that need to be addressed?

The answers to this question are supplemented in Conclusion section (colour highlighted).

The integration of the PEMFC monitoring system with other onboard systems is possible. However, its complexity will also depend on the functionalities of the PEMFC controller (HFCT), that controls the operation and monitors key parameters of the H60 as shown in Figure 3. Depending on the available documentation from the HFCT manufacturer, the monitoring system would be able to monitor the operation of the PEMFC controller and simultaneously monitor the power ratios at its input, output, and on-board power supply of the UGV.

The possibility of intervening in the controller software is also key to expanding the functionality of the monitoring system in relation to monitoring the PEMFC H60. If necessary, a GPS module (e.g. NEO-6M), which is compatible with most Arduino development platforms, can be integrated into the monitoring system.

 

Author Response File: Author Response.pdf

Reviewer 4 Report

Comments and Suggestions for Authors

The study contributes to the development of sustainable energy solutions for mobile platforms, highlighting the practical challenges and efficiencies of PEMFC systems. The integration of wireless monitoring enhances real-time diagnostics and optimization potential. There are some comments and suggestions to improve the article a little further. Overall, this paper requires minor revisions.

1.Line 141-152: Missing formula 9.

2.Line 261: The first letter of the title is not capitalized, which is inconsistent with the title format of other chapters.

3.Line 387: In Figure 9(b), the position of the caption blocks the picture information. It is recommended to adjust the position of the caption.

4.Line 498: Figure 22 shows the transmission data from HC-05 Bluetooth module to smartphone. Displaying these data has little effect on this article. It is suggested that Figure 22 be deleted.

5.Line 539: Discussion is chapter 5, conclusion is chapter 7, chapter 6 is missing.

Comments on the Quality of English Language

The English could be improved to more clearly express the research.

Author Response

Reviewer 3 (magenta coloured answers and corrections)

Dear Mr./Ms./Mrs.

thank you very much for your review. We have try to answer/correct our article according to your reccomendations.

Please see text bellow and attached pdf file. In attached pdf file, the magenta higlighted text are our corrections according to your recommendations.

Please, the rest of highlighted text/figures are answers to others reviewers.

Thank you again for your time

M.

 

1.Line 141-152: Missing formula 9.

Mistaken equations numbering was corrected ... (eq. 8-9)... Please see attached pdf file, line 138 and 145.

2.Line 261: The first letter of the title is not capitalized, which is inconsistent with the title format of other chapters.

Corrected according to your requirements. Please see attached pdf file, line 299.

Line 387: In Figure 9(b), the position of the caption blocks the picture information. It is recommended to adjust the position of the caption.

Corrected according to your requirements. Please see attached pdf file, line 445-446.

4.Line 498: Figure 22 shows the transmission data from HC-05 Bluetooth module to smartphone. Displaying these data has little effect on this article. It is suggested that Figure 22 be deleted.

Figure 22 was deleted according to your requirements.

5.Line 539: Discussion is chapter 5, conclusion is chapter 7, chapter 6 is missing.

Corrected according to your requirements - please see attached pdf file.

Author Response File: Author Response.pdf