Thermodynamics-Based Predictive Maintenance for Medical Devices

Because every mechanical device wears down eventually, predictive maintenance is crucial for the life cycle of medical devices. In this article, Nikolay Khabarov, Principal IoT/IoMT Architect, describes how thermodynamics can be used to create a digital twin of a mechanical part of a medical device to monitor the device’s performance.
5 min read
08/06/21
By Nikolay Khabarov
Principal IoT/IoMT Architect
Share
Thermodynamics-Based Predictive Maintenance for Medical Devices

Reliability is essential for mechanical medical devices. Unlike computers and other digital devices that are pretty reliable, mechanical devices are prone to breakdowns. Industrial equipment downtime can be costly. When it comes to the Healthcare and Life Sciences industry, medical equipment breakdowns may cost more than money — potentially leading to inaccurate diagnostics results, worse patient outcomes and longer hospital stays. That is why medical device maintenance is critical for healthcare systems. Proper maintenance can help you improve patient care and safety, reduce risks, and save money. 

Fortunately, modern digital technology can help predict future breakdowns, provide proper maintenance, and avoid serious medical device failures. Let’s take a look at possible solutions that can be employed by healthcare systems and medical device manufacturers. 

Classical Approach to Predictive Maintenance

Medical devices are in fact physical objects containing some mechanical parts with attached sensors that collect and analyze data about the condition of these parts. For example, simple temperature sensors are installed on a motor, solenoid, actuator or gearbox of a mechanical device. When the temperature exceeds the threshold, the system signals that something is wrong. In some devices, we can also install modern MEMS (micro-electro-mechanical systems) accelerometers to measure vibration level. Likewise, when the vibration level is higher than a predetermined limit, the device’s motor is likely broken or operating incorrectly. In these scenarios, deviations of temperature and vibration levels indicate that something is already wrong and the device is not functioning properly.  

Since these breakdowns can be costly, it is useful to be able to predict malfunctions before they occur. Machine learning algorithms (ML) can be employed to identify a problem before a device is out of order. While ML is quite effective in digital technology, prediction model development is a challenge. Additionally, these models require training using real data from actual devices. This means that we need a huge dataset with sample data gathered under different conditions and from different use cases. As a result, creating ML predictive maintenance models can become complicated and costly.

Digital Twins as an Option

Digital twins are a different approach, which involves creating models that are mathematical simulations of objects or processes. A model does not have to be an exact copy of a physical object. Instead, it describes specific properties of this object. For example, you can easily calculate the location of a car that travels from point A to point B at a constant speed. All you need to do is multiply the car’s speed by the time it takes to go from one point to the other — there is no need to account for details like the make or model. This simple math demonstrates how an abstract model of a moving object can be applied to a particular car without knowing additional details. 

The same principle works for predictive maintenance. Thermodynamics can be used to create a digital twin of any mechanical part of a medical device. From there, we can estimate its condition. Laws of thermodynamics are fundamental, and it is easy to build a model that shows how an object’s temperature changes. All we need to know is the amount of heat that this mechanical part produces, and we can calculate the temperature of that part.

We know the amount of power consumption for motors, solenoids or any mechanical actuators. Every device involves parasitic work, which is mostly due to heat they produce when operating. The ratio of useful work to parasitic work indicates the efficiency of the device. A lack of efficiency is the first sign that something is wrong. To estimate efficiency, we apply thermodynamics laws of physics and calculate changes in temperature when the device heats up. 

To illustrate the idea, let’s run an experiment with a wrist blood pressure monitor motor.

The blood pressure monitor has a motor that pumps air. We will attach a thermo sensor to the motor to measure its temperature. For the experiment, we measure the temperature when the motor is operating at its full power. You can see the results of the measurements in the graph below:

Graph 1

Graph 1

We then record the mechanical device specification to establish how the temperature is supposed to change when the motor is operating at its full power. When we put two graphs on top of each other (see graph below), we can see how much the motor deviates from the expected outcome. 

Graph 2

Graph 2

Now we have a simple mathematical model of how the device’s temperature should change when the power is on. When something is wrong, the temperature will grow differently. A tiny temperature deviation tells us that our device is becoming less efficient and is not operating properly. On the graph below, you can see a stronger deviation from the normal when the motor produces 20% more heat than it is supposed to. This indicates that something is wrong and the device requires immediate attention.

Graph 3

Graph 3

Conclusion

The approach described above is based on fundamental principles of physics and is easy to implement. In fact, it can be applied to any medical device that consumes electricity. As you can see in sample graphs above, it is easy to detect temperature deviations. This simple experiment with a wrist blood pressure monitor demonstrates the capabilities of this approach, which can be applied to more complicated medical devices and laboratory equipment. The solution can be useful when it comes to building a robust and stable device that automatically reports its status and maintenance requirements. 

Sign Up for Updates!

Subscribe now to receive industry-related articles and updates

Choose industries of interest
Thank You for Joining!

You will receive regular updates based on your interests. No spam guaranteed

Add another email address
READ MORE
Sign Up for Updates!
Choose industries of interest
Thank You for Joining!

You will receive regular updates based on your interests. No spam guaranteed

Add another email address
Welcome
We are glad you found us
Please explore our services and find out how we can support your business goals.
Get in Touch Envelope