Hello, I recently published an article about creating a naval vessel marine engine application with LightningChart .NET. Such article raised a good interest in the topic within the LinkedIn community and decided to write a follow-up article about the most important questions I received about it.

Even though I’m not an expert on Naval Vessels, I needed to simulate situations where the condition of crucial components in complex systems was being measured. That’s when the idea came to me to research Naval Vessel’s Marine Engines, learn more about them, and at the same time learn about a topic that is unfamiliar to me. 

I researched and read some information about key components in such vessels. I’ve previously worked in oil, food, and other industries as an industrial air compressor maintenance technician, so I have some knowledge of similar or slightly related components. 

Below are the points I decided to include in the monitoring dashboard and the reasons behind them: 

1. Turbine Temperature: Temperature is one of the most important parameters because it indicates whether the turbine is operating within its safe range. Excessive temperatures can damage the engine’s internal parts and reduce its lifespan. It also helps prevent overheating, which could lead to catastrophic failures.

2. Fuel Flow: Fuel flow regulates the amount of energy the gas generator receives. Monitoring it ensures that the engine receives enough fuel to operate at its optimal power and efficiency level. Insufficient fuel flow could result in malfunction, reduced efficiency, or even system failures.

3. Turbine Injector Control: The turbine injectors control the amount and mixture of fuel entering the combustion chamber. If they are not functioning properly, it could cause poor combustion, reducing efficiency and negatively impacting the generator’s performance. Additionally, poor injector control could lead to excessive pollutant emissions.

4. Gas Generator RPM (GGn): RPM indicates the rotational speed of the turbine and the generator. Maintaining a constant speed within the optimal range is crucial for system stability and to avoid damage from overspeed or underspeed, which can occur if not properly monitored.

5. Turbine Engine Thrust Torque (GTT): Thrust torque is the measure of the force generated by the turbine to move the generator. Monitoring it helps ensure the turbine is producing the right amount of power for the generator to operate efficiently. A low thrust torque may indicate mechanical issues, such as excessive friction or wear in the turbine component. 

Unfortunately, this data was obtained from a public dataset on the internet, but I am currently working on the development of electronic monitoring modules for a glass manufacturing plant in my city. The process is quite personal, so I might be wrong, but I hope it is useful to you. 

It is worth mentioning that the dashboard developed in .NET uses data processing logic that could be useful to you at a more advanced stage once you obtain sensor values. 

Sensor Assembly

At this point, it is necessary to analyze which sensors can withstand the environmental conditions (temperature, surface, dust levels, humidity, etc.). Let me mention the types of sensors you could use for each component:

• Turbine Temperature – Temperature Sensors (Thermocouples or RTDs) : Thermocouples are common sensors for measuring high temperatures in turbines. They can withstand high temperatures and provide accurate readings. RTDs (Resistive Temperature Detectors) are also used in some applications, as they are more precise within a moderate temperature range but are less common in gas turbines. 

• Fuel Flow – Mass or Volumetric Flow Sensors : Flow meters such as ultrasonic flow meters, electromagnetic flow meters, or Coriolis mass flow meters are used. These sensors can measure the amount of fuel passing through pipes, either in terms of volume or mass, which is key for controlling the amount of fuel entering the turbine. 

• Turbine Injector Control – Pressure and Temperature Sensors on the Injector :To monitor the functioning of the fuel injectors, pressure and temperature sensors are used in the injection system. These sensors help ensure that fuel is injected correctly and under ideal conditions for proper combustion. 
• Gas Generator RPM (Revolutions per minute) Speed sensors (Encoders or Tachometers) : Rotational encoders or tachometers are used to measure the rotational speed of the turbine shaft. These sensors provide precise RPM readings and allow control of the generator speed, preventing overspeed or under speed.
   
• Turbine Engine Thrust Torque (GTT)- Torque Sensors (Torque Transducers) :To measure the thrust torque, torque sensors are used, which are designed to measure the rotational force generated by the turbine. These sensors are typically mounted on the turbine shaft and can measure both torque and the generated power. 

As an important note, I assume that the temperature, kinetic ranges, and other factors are much higher than what we would normally sense in daily life. Therefore, you would need to research with industrial sensor manufacturers which models could offer you sensors that can withstand such conditions.

As a recommendation, you could work with smaller components that allow you to experiment. I would recommend using more accessible sensors that can help you develop your programming logic. Many industrial sensors share similar logic to microcontrollers such as the ESP 32 and Arduino. Therefore, your development wouldn’t be far off from what would be a real-world practice (except for assembly pieces and dynamic ranges). 

Once you have the sensors, depending on your budget, you can choose from various data processing options. 

Embedded Sensors:  These sensors work as independent modules, where within a small casing (steel box, structure, or other materials) you will find the sensor, a microprocessor (e.g., ESP 32, ARDUINO), and electrical components (wires, resistors, batteries, etc.).

These modules aim not to rely on a computer and can sense on their own once powered.  How the data is collected and transmitted will depend on your programming logic added to the microprocessor. Some sensors do not require a microprocessor because they already come with one integrated, so the first example applies to embedded sensors that you create yourself. 

Integrated Sensors:  As I mentioned earlier, these sensors do not come with a built-in microcontroller and require an external microprocessor, which (at this point) will be in a remote location, and communication is carried out via wiring, Wi-Fi, or Bluetooth. 

The microprocessor will be responsible for receiving the sensed data, and the logic you generate will create an active response (e.g., a JSON Blob), which you can use to identify each parameter and use it later. For example, the ESP 32 and Arduino can be programmed through the ESP or Arduino IDE, which works perfectly on a PC. 

In your practical case, I would recommend that you assemble or obtain embedded sensors, as the distances will be very long, and in the case of underwater sensors, wiring will be a very complicated task. You will need to use long-lasting batteries or implement energy generation mechanisms (e.g., a dynamo). I would also place Bluetooth or Wi-Fi antennas with wiring to a safe transmission area so you can collect this data. 

At this point, you will have several options to obtain and store data from your sensors. You will need a server, a central computer that receives the signals from your sensors. Each sensor will be a port, and you will need access to each port to consume its active responses.  

For example, C# offers the System.IO.Ports tool, which allows you to access a specific port and get the active response from the sensor (e.g., JSON Blob). Once you have obtained that response, you will need to process these responses and store them in a database, flat file, or any other source you prefer. 

If you require the monitoring to be visible on the web, you will need to implement a web socket, which is like a web service that receives responses and HTTP requests, but in a more agile way than a traditional web service. A web socket will always be active and will receive data second by second. This web socket should be on the web server. 

Once your infrastructure is set up, the data transmission speed and activity periods will depend on the quality and power of the sensors, the microcontroller programming, the server’s power, and the quality of your code. You can schedule tasks that will run your web socket or API at specific hours and/or days. 

Do you set alarms or thresholds with these parameters to flag irregularities?  Yes, a fundamental part of monitoring is to inform about issues or irregularities in the components. Let’s say your turbine has a maximum temperature of 150°C and a minimum of 80°C. You should set this temperature range in your API.  Any value above or below this range should trigger an alarm. What I do is, using C#, alert via email and UI (dashboard) to the engineers or supervisors. You can implement an alarm that triggers via PLCs when your computer receives a dangerous signal. 

Can your program predict potential failures and generate trends based on the collected measurements? The dashboard in this article is quite basic, and its purpose is more educational. However, in a more real-world development, you can certainly implement algorithms based on the collected data. As I mentioned earlier, based on the technical sheet of each component, you should set ranges. I think the easiest part is detecting numerical values, the way you alert will be the most complex part. Implementing alarm systems, email message sending, mobile apps, or other options will be the real challenge. 

Have you explored similar monitoring systems for hull and propeller performance? Unfortunately, I haven’t delved into that topic, but I could do so. 

I’m not an expert in electricity, but I can share a real case. Currently, in my development, we don’t need to worry much about this issue since the plant has its own battery backup system. If the city power goes out, these batteries activate, with enough autonomy to last for a day. Now, on a more local level, the servers where your data is hosted should have protection systems for power fluctuations, such as uninterruptible power supplies (UPS). These will help keep your servers stable if the local power source switches to battery power. 

Is it able to suggest adjustments for optimizing injection or adjustments to improve fuel efficiency?

I believe this area is more suited for other specialists. I suppose that, with historical levels of temperature, vibration, noise, etc., a mechanical engineer and even a chemical engineer could recommend the use or replacement of mechanical and chemical components (lubricants, for example). This task requires analyzing various factors, from lubrication and fuel type to the environment surrounding the component. 

Have you incorporated any parameters related to air pollution, and if so, how are they measured?

I have used some sensors to measure the level of gases in the environment, alongside temperature sensors. There are gas sensors, particulate matter sensors, volatile organic compounds, electrochemical sensors, optical sensors (laser or scattered light), NDIR sensors (non- dispersive infrared), and metal-oxide semiconductor (MOS) sensors. I haven’t used any of these, as many are difficult to obtain or are expensive. I researched some models that you might want to check out, which could help you in more detail:

• Electrochemical: Alphasense (CO, NO2, O3, SO2), Figaro TGS-5340 (VOCs, H2). 
• Laser/Optical: Plantower PMS5003 (PM2.5, PM10), Honeywell HPMA115S0 (PM2.5, PM10). 
• NDIR (Non-dispersive Infrared): MH-Z19 (CO2), Senseair S8 (CO2), K-30 (CO2). 
• MOS (Metal-oxide Semiconductors): Figaro TGS Series (various gases and VOCs), CCS811 (VOCs, CO2).

In summary, most of the work in sensing will be in the implementation of the IT infrastructure, whether it’s setting up a server, developing a monitoring environment, or processing data. Of course, installing sensors can be complex or dangerous depending on the environment in which they are placed. I hope my experience and ideas can help you. My recommendation is to start with a medium-sized prototype that allows you to experiment with low-cost components.

This way, you can develop processes that can withstand communication interruptions, power failures, and sensor malfunctions.  Try experimenting with signal quality in different environments.  Make sure that the signal from your sensors is not affected by wind, noise, or vibrations. Try to use modern programming frameworks, which are increasingly optimized and powerful, making your development much easier. Thank you!