Team Status Report 3/8

What are the most significant risks that could jeopardize the success of the project? How are these risks being managed? What contingency plans are ready?

The most significant risk to our project is that the suction-based end effector won’t be able to pick up different types of trash items. We’re managing this risk by focusing a considerable amount of time on the development of the end effector mechanism. In the case that the suction with the vacuum pump is ineffective, we plan to pivot to a claw like end effector mechanism for picking up the pieces of trash.

Were any changes made to the existing design of the system (requirements, block diagram, system spec, etc)? Why was this change necessary, what costs does the change incur, and how will these costs be mitigated going forward?

There are no current changes made to the existing design of the system.

Provide an updated schedule if changes have occurred.

There are no schedule changes as of now.

Part A was written by …., Part B was written by Ethan, and Part C was written by Alejandro.

Part A:
with consideration of global factors. Global factors are world-wide contexts and factors, rather than only local ones. They do not necessarily represent geographic concerns. Global factors do not need to concern every single person in the entire world. Rather, these factors affect people outside of Pittsburgh, or those who are not in an academic environment, or those who are not technologically savvy, etc.

Currently, the issue of proper waste processing is one that many countries, including the US, are struggling with. Many countries do not have the infrastructure to pay for recycling facilities, as it requires a lot of manpower, and the resulting materials are not very profitable. Additionally, a lot of our waste is shipped to other countries in order to prevent it from piling up here, however this means that other countries are burdened with our trash. SortBot could help reduce the amount of trash by separating out the useful materials, at a much lower cost compared human workers. This could potentially help other countries without the funds for recycling management to do so.

Part B:

In the United States, the common practice is to dispose of all types of municipal solid waste in single bin, prioritizing convenience over environmental concern. SortBot is set out to challenge this norm, one that focuses on ease and efficiency, by introducing an autonomous system that streamlines waste separation without requiring a behavioral change. By shifting the responsibility of sorting waste away from the person throwing away the trash, SortBot can work within the existing American cultural norms. This ensures that materials can be property sorted without requiring individuals to change their habits. While this is not the best solution for this issue, it is one of the most appropriate for current American culture.

Part C:

Our product solution will meet the need of the consideration of environmental factors by providing a streamlined method to making the environment cleaner. Utilizing advanced computer vision and machine learning algorithms, our robot has the ability to identify and differentiate between different types of trash and can thus correctly categorize the trash types and sort them into their respective bins. The system can aid in the sorting of trash to aid in the recycling of these items. Normally, all recycling items mixed with trash would normally end up in landfills and be lost in terms of the value that they could provide when recycled, but now with our product solution, we can recover these items and help in reducing the environmental strain of waste disposal.

Alejandro’s Status Report for 3/1

I spent the majority of the week working on the assembly of the x-y axis of the gantry system. The assembly was time-consuming and laborious. The assembly requires a few more steps that will take some considerable amount of time. I will work on finishing the assembly the week I get back from Spring break. My progress is slightly behind schedule. In this case, I will prioritize the assembly of the gantry’s xy axis in the following week of work. Additionally, I was experimenting with the use of the canvas drawing onto the camera feed but I will need more time to solidify the feature. I’m also running a bit behind with the canvas drawing feature but I’ll have that done promptly after assembling the gantry. So next week I intend on completing the gantry xy axis and finishing the implementation of the bounding box HTML feature of mine.

Ethan’s Status Report for 3/1

The majority of effort this week was spent on finding bugs in our YOLO codebase that caused the results from last week to look really poor. I discovered it was the loss function. Previously I was using a more naive approach that combined a weighted mean-square loss for the bounding boxes with a weighted cross-entropy loss for classification. After reading a couple of Medium articles about YOLO, I realized that I was implementing an entirely different loss function. Once I fixed that I also a little bit more training infrastructure that would hopefully making training analysis easier: curve plotting for each loss. By monitoring loss, I can identify parts of the model that would need tuning in future runs.

Next week, I plan on getting detection working on toy images.

Currently I am on schedule.

Teddy’s Status Report for 3/1

Over the last two weeks, I worked on constructing the gantry as well as the frame that holds it. I was able to construct the 4xidraw part of the gantry since all of our parts had finished arriving. I was then able to print, assemble, and attach the z-axis to the gantry. Currently, the parts required for the z-axis movement are assembled, but the end-effector is not attached, since we are still waiting for the vacuum pump and the relays to arrive. There is an issue in that the z-axis can’t extend very far away from the gantry, as the weight causes the rods to sag. In order to remedy this, I’m planning on attaching a bar horizontally onto the part where the z-axis attaches to the rest of the gantry, which will rest on two other bars so that the front does not sag. A picture of the current gantry is shown below.

I am a little behind schedule, partly due to delays in the shipping of the gantry parts as well as the end-effector parts. I plan to fix the sagging issue on the gantry next week, as well as hopefully get the electronics and the code for the 4xidraw working.