Posted on

Mahati’s Status Report for 04/25

This week, I focused on improving our robot’s data collection and classification pipeline. I worked closely with Rayann to continue training the robot’s SVM model, refining its ability to distinguish between defective and non-defective regions. As part of this effort, we also re-evaluated our feature set to identify the most informative signals, aiming to enhance model performance and reduce noise in the data.

On the hardware side, we made significant progress by assembling the 3D-printed components of the robot, bringing us closer to a fully integrated system. We also added a second acoustic emission sensor, allowing us to collect data more efficiently and speed up the testing process. This improvement is especially valuable as we scale up data collection for training and validation.

Additionally, we revisited our data strategy and made an important decision regarding dataset composition. After analyzing the differences between handheld and robot-collected data, we observed that the sensing characteristics varied significantly. To maintain consistency and improve model reliability, we decided to exclude the handheld data and focus solely on robot-collected samples for training moving forward.

Posted on

Mahati’s Status Report for 04/18

Our teensy mcu broke, so I rewired the teensy mcu with the ADC, motors, and sensors. This time, while rewiring, I made sure to be very thorough and neat with the way that I was wiring so that it would not break again. We have been very careful storing the parts as well, so that there would be no issues for our final. 

I also worked alongside Rayann to test and train our robot’s SVM model, helping ensure that our classification pipeline was functioning correctly. We tested the model with the robot and added robot data to the training dataset. We collected 200 raw voltage data values, processed the features, added them to a robot dataset, and then let the robot move forward through the duct for 1 second. 

As you’ve designed, implemented, and debugged your project, what new tools or new knowledge did you find it necessary to learn to be able to accomplish these tasks? What learning strategies did you use to acquire this new knowledge?

I had to learn several new tools and technical skills. I gained experience booting and setting up an NVIDIA Jetson Orin Nano, writing Arduino code, and wiring a Teensy microcontroller with ADCs, motors, and sensors. As I am more on the software track, I sought out help and did the wiring alongside Adithi, and that was very helpful.

To build this knowledge, I used online tutorials and YouTube videos to quickly get hands-on familiarity with the hardware and software tools. Since I was working on soldering in AirLabs, I got to ask the researchers there about their inputs and was helped by Mark. This helped me effectively troubleshoot issues and integrate different components into a working system.

Posted on

Mahati’s Status Report for 4/4

This week’s progress:

This week, I focused on refining the machine learning model, with the goal of improving both its accuracy and overall robustness. I explored ways to better structure the data and tune the model so that it can generalize more effectively to real-world scenarios. This refinement process is an important step toward ensuring that the model performs reliably when integrated into the full system.

In addition to the machine learning work, I made progress on the Jetson-side implementation. I restructured the codebase by separating the robot control logic from the ADC signal processing pipeline. This separation makes the system more modular and significantly improves the ability to visualize and debug sensor data independently from control behavior, which will be valuable as the system becomes more complicated.

Things to do for next week:

Looking ahead to next week, I plan to collect more training data to further strengthen the model. With a larger and more diverse dataset, I aim to improve the model’s performance and reduce potential edge-case errors. I will also continue testing the system end-to-end to ensure stable and consistent communication between all components, including the Jetson, Teensy, and sensors.

Additionally, I will be collaborating with Aditi to integrate the LiDAR system and begin working toward autonomous navigation of the robot through the HVAC environment. This will involve combining perception and control components into a more cohesive system. 

My work can be found here: https://github.com/aphadke234/ece_capstone_C7

Posted on

Team Status Report for 03/21

This week, our team focused on progressing several core hardware and embedded system components required for our experimental setup, including motor control, ultrasonic sensor testing, and preparation for the solenoid tapper system.

Aditi worked on developing the embedded motor control logic for the robot by implementing and testing the control signals used to drive the motors through the microcontroller and motor drivers. This work focused on ensuring that the motors respond consistently to control inputs and that the embedded code interfaces correctly with the hardware so the robot can operate reliably during experiments. At the same time, Rayann worked on characterizing the ultrasonic sensing system that will be used to collect acoustic data from HVAC duct samples. This involved testing how the ultrasonic transmitter and receiver respond to different input voltages. In addition, Mahati continued preparing the solenoid tapper control system, which will eventually generate consistent mechanical taps on the duct surface, by reviewing and iterating on preliminary control code so it will be ready for integration once the ADC and data acquisition pipeline are in place.

Overall, the team made progress toward getting the robot and sensing system ready for integration and future data collection. While we are slightly behind schedule due to the ADC not arriving on time, we feel confident that we can catch up once the remaining hardware is in place. Unfortunately without the ADC, we cannot fully record the sensor measurements for our algorithm.

Next week will be focused on completing hardware integration and beginning the process of collecting experimental data so that we can start testing our algorithm. Since we are approaching a two-week milestone, the primary goal will be ensuring that the robot and sensing pipeline are fully functional. One of the most important tasks will be beginning structured data collection from the duct samples. Once the ADC arrives and is integrated into the system, we will start recording acoustic signals that can be used as input for our algorithm. Having real testing data will allow us to evaluate our signal processing pipeline and begin running the classification algorithm on actual measurements rather than simulated or preliminary signals.

Another major priority will be getting the robot fully operational. This includes ensuring that the motors, sensors, and control systems all work together reliably. The team will focus on integrating the embedded motor control, sensor inputs, and solenoid tapper so that actuation and data acquisition can occur in a synchronized and repeatable way. Additional work will include receiving and inventorying the remaining hardware components, integrating the ADC with the sensing system, finalizing the solenoid tapper hardware integration, and verifying that the full pipeline operates correctly.

Potential issues/risks include further delays in ADC delivery or malfunction, synchronization issues between the motors, solenoid, and sensors, or unexpected behavior in the ultrasonic sensor readings on different duct materials. Despite these risks, we feel that our progress so far has been solid. We have made significant headway on motor control, sensor characterization, and solenoid preparation, and with the remaining hardware arriving soon, we expect to meet our two-week milestone and be prepared for the interim demo.

Posted on

Mahati’s Status Report for 3/21

This week’s progress:

  1. Developing the Control Code for the Solenoid Tapper
    This week, I worked on writing and testing the control code for the solenoid tapper that will be used to generate consistent mechanical taps during our experiments. The goal of this component is to produce repeatable impulses on the HVAC duct surface so that we can reliably capture acoustic and vibration responses from the system. I focused on implementing the basic control logic needed to trigger the solenoid and regulate the timing between taps. This would involve experimenting with different timing intervals and pulse durations to ensure that the solenoid actuates reliably and returns to its resting position between activations. I got the preliminary code started a few weeks back and have been iterating on that, so when we have the ADC, we will be able to properly capture the data needed with the solenoid tapper. 

My work for the week can be found here:
https://docs.google.com/document/d/1qRRPVxLNIO-z_C0ETqJalA6X_706GjrcQP8EjuErZ6U/edit?usp=sharing 

Things to do for next week:
Next week, most of the hardware materials for the experimental setup should arrive, which will allow us to begin assembling the system and collecting preliminary data. The main tasks for next week include:

  • Receive and inventory all incoming materials required for the experimental setup. We are specifically waiting for the ADC
  • Get the motors to work 
  • Finalize the solenoid tapper hardware integration with the rest of the system.
  • Begin collecting initial acoustic and vibration data from the duct samples.
  • Set up and test sensors with the data acquisition hardware to ensure signals are captured correctly.
  • Work with Adithi to integrate the solenoid control logic with the broader embedded system so that tapping and data collection can occur in a synchronized and repeatable manner.
Posted on

Mahati’s Status Report for 3/14

This week’s progress:
This week, I focused on understanding the physical assembly of the HVAC duct structures that we will be using for data collection. We received several square metal duct pieces from the FMS shop, but when we attempted to cut and assemble them, it turned out to be more complicated than expected. The ducts are held together using riveted sheet metal joints, and it was initially unclear whether we should remove the rivets, re-rivet the pieces ourselves, or modify them in another way. We tried to cut a part off of the HVAC ducts; however, just the process of getting the HVAC duct cut took 1 hour, and we have 20 more to go, and that’s not including putting the ducts back together. 

To address this, I spent some time researching how HVAC sheet metal ducts are typically assembled and how we might work with the pieces we currently have. In parallel, Professor Ed from FMS shared the contact information for fixit@andrew.cmu.edu, which handles fabrication and facilities requests. I reached out to them to see whether it would be possible to obtain additional duct sections that are already pre-connected or easier to work with for testing, and I am currently waiting for a response. If we can obtain pre-assembled ducts, it would significantly simplify the experimental setup and allow us to focus more on data collection rather than mechanical assembly.

In addition to the mechanical work, I began drafting the preliminary code for controlling the solenoid tapper that will be used to excite the ducts during testing. The solenoid will act as a consistent impact mechanism so that we can generate repeatable acoustic and vibration signals from the duct surface. While working on this, I also reviewed and worked to understand Aditi’s motor control code so that we can eventually integrate the actuation components with the rest of the robot’s system.

Finally, I spent time in the lab working with Aditi and Rayann to assemble most of the robot platform. We were able to get the majority of the structural components in place and ensure that the main hardware elements fit together properly.

My weekly progress (research for HVAC duct methods and code): https://docs.google.com/document/d/16fPUd4hhc9Xqjzm7Y71jKUkrX2CyWH6wOT2W3ph-IuI/edit?usp=sharing 

Things to do for next week:
Next week, the focus will be on moving from assembly and planning into more integrated system testing and data preparation.

One of the main goals will be to create a structured training and data-collection plan for the remainder of the project timeline. This will outline how many recordings we want to collect per duct condition, how the data will be labeled, and how we will structure the dataset so that it can be used effectively for model training.

In parallel, I will work on getting the robot fully operational. This includes integrating the motor control code, the solenoid tapping mechanism, and all of the sensors so that they can operate together within a single system. The goal is to reach a point where the robot can move along the duct, trigger the solenoid tapper, and collect synchronized sensor data automatically.