Posted on

Mahati’s Status Report for 2/7

Since I will be working primarily on the software component of the project, I wanted to explore testing approaches beyond static inspection datasets. Ideally, we would like a way to generate signals and simulate sensor feedback so we can observe how the robot responds in near real-time, closer to how it would behave in a real deployment.

To this end, I am investigating whether there are HVAC duct simulation tools that could (1) imitate the duct environment and (2) generate realistic signals or environmental conditions for the robot to interact with. So far, I have identified several duct-specific simulators, including SimScale, TensorHVAC-Pro, and OpenFOAM. Over the next few days, I plan to explore these tools in more detail and evaluate whether they are feasible to run on my laptop, since compatibility and setup overhead can be a limiting factor.

If these simulators do not meet our requirements—particularly for real-time interaction or sensor-level simulation, I will pivot to more general robotics simulation frameworks such as NVIDIA Isaac Sim, Webots, and Gazebo. These tools are not duct-specific but offer better support for robot dynamics, sensors, and closed-loop testing. Since I have limited experience working with simulators outside of ROS, I expect this to be a learning curve over the coming week.

Real-world hardware testing was also a major concern raised during project proposal feedback. I contacted the Pittsburgh Trade School to ask whether they had unused duct components we could use for testing, but I have not yet received a response. I plan to follow up via email and reach out to additional trade schools in the Pittsburgh area over the coming week to see whether sourcing duct parts for physical testing would be possible.

Summary

Things accomplished this week

  1. Identified both duct-specific and general-purpose simulators relevant to the project.
  2. Goals for next week
  3. Install and evaluate selected simulators, focusing on compatibility and real-time sensor simulation capabilities.
  4. Follow up with the Pittsburgh Trade School and contact additional trade schools to explore acquiring unused duct components for testing.

My week’s work can be found here: https://docs.google.com/document/d/1MUVXOmcYyLjaods_Zd165aJV-86cpB4vcz8pvJeTxqY/edit?usp=sharing 

Posted on

Adithi’s Status Report for 2/7

Accomplishments:
This week, I focused on researching the hardware requirements for our onboard intelligence. I read through multiple papers and machine learning datasets to evaluate which public, free datasets we could use for our model’s initial training. I quickly learned it is hard to find a lot of acoustic datasets for our use case but fortunately I came across academic and industrial research work related to Structural Health Monitoring. We can use guided wave-based structural monitoring datasets to evaluate how taps against sheet metal in HVAC ducts can propagate through wave materials. I was able to identify how frequency bands and wave reflections can indicate leaks or corrosion. In all, I believe I have found 3 or more datasets we can use to train our model. I have shared these with the signal processing expert in our group, Rayann, and it is pending discussion.

On the hardware side, I further analyzed whether the NVIDIA Jetson Orin Nano is the right onboard computing platform for us to run real-time signal processing and ML interference on. Upon further research I was able to look at performance, power requirements, and software support as well as ability.

In addition to the Jetson, I looked into the best tapping mechanism for our design. I found 3 but have narrowed it down to solenoid, and servo-driven approaches as this best fits our use case and how we are planning to build our robot.

I also, along with my teammates, contacted several HVAC repair places and learnt that it is best to contact trade schools, dump sites, and look within CMU itself in the robotics department and the Mechanical Engineering department. I will be doing this early next week as we need some way to source sheet metal and learn more about the defects themselves in general to be able to simulate realistic testing towards the end of our build process to ensure our product works.

Finally, I ordered most of our initial parts for our project and am excited to start building next week!

My work is here: Adithi – Week 4

Schedule:
We are slightly behind schedule because we have been researching technical components and restrictions of the testing of our project. However, because of our research, I think we are putting ourselves in a better position by addressing key problem areas and thinking about how we can mitigate them beforehand. We will be catching up to our schedule by Wednesday latest this following week.

Next Week:
Next week, I will finalize the hardware architecture, component selection, and finish my research on the power distribution and mechanical integration. I will also start building the tapping mechanism if the parts arrive otherwise, I will start finalizing testing and sourcing plans for different materials and begin the software development with Mahati. 

Posted on

Rayann’s Status Report for 2/7

This week, I presented our Proposal. I spent last weekend sorting out the elements we wanted to include with my team, specifically the use case requirements. I created some of the figures and worked on the technical challenges, solution approaches, and testing, verification, and metrics slides. I also practiced and revised how I should present and how I should answer questions. We were able to get a clear picture of the project we want to accomplish this semester.

I also looked into locating HVAC ducts for physical testing. I called a couple of HVAC duct repair and replace companies and inquired about where they store damaged ducts, but they were focused on business, i.e. they wanted me to get my ducts repaired even though I explained that I have no ducts. We, as a team, are continuing our search for used ducts but as a backup, I have found HVAC ducts of different sizes to buy on Amazon. I have also found fiberglass board to buy on Amazon to apply to our sheet metal ducts so that we are able to test our system on fiberglass duct board. Fiberglass duct board consists of standard sheet metal ducts that have a fiberglass lining that serves as an extra layer of insulation

I looked extensively into acoustic features that indicate corrosion and leaks. Although I had a general understanding of the type of acoustic signature that indicates these defects, I was able to find more specific information as well as some signal processing techniques to employ from research papers. These papers were done on a variety of sheet metal materials, but I was not able to find research papers on the acoustic signatures of defects on fiberglass duct board. Fiberglass duct board helps to dampen noise. My worry is that it will be difficult to generalize the ML model to fiberglass duct board because the acoustic signature that we capture with the robot will be dampened, and it may be hard to extract the acoustic features from the data. I am also worried about the conversation my team and our supervising professor have been having regarding the data to train the ML model. We had previously assumed that we could find a dataset with labeled acoustic data from the duct materials of our choice, but this is proving difficult. Our professor has recommended that we gather our own data based on our robot’s specific ability to create and collect acoustic data. Currently, we are working on getting our robot together while looking at potential datasets. Whichever dataset we use must, at the very least, include acoustic data labeled with the defects we are focusing on (corrosion and air leaks) from the duct material we are focusing on (sheet metal and fiberglass duct board). 

Next Week Goals:

Looking through datasets to choose one that meets the minimum requirements stated above

Finding a viable HVAC duct source and purchasing/obtaining the ducts

Developing the design of the acoustic data processing pipeline

My work for this week can be found here: https://docs.google.com/document/d/1RLk7pQB5754Au6vyzFaVA-jm3ABx_Y89l1xmLTPi_bc/edit?tab=t.0