Skip to content

EECS 441 - 443 Project folder with files for senior capstone project for Jorian Bruslind, Mack Hall, and Zach Bendt

License

Notifications You must be signed in to change notification settings

Jbruslind/ECE44x_Senior_Design

Repository files navigation

ECE44x_Senior_Design

ECE 441 - 443 Project folder with files for senior capstone project for Jorian Bruslind, Mack Hall, and Zach Bendt

Brief descriptions of each member's responsibilities are detailed here. Below is a list of main projects for each team:

Automated Microbial Analysis

The Automated Microbial Analysis project aims to develop a system which is able to analyze a series of microbial samples on a special media called PetriFilm automatically. PetriFilm is a 3M product that is used in a variety of scientific capacities, mainly in food science and the medical field.

PetriFilm is a very thin paper-like substance which has 2 layers: on one there is an agar media film that is used to grow bacterium, on the other there is a wax coating that is used to seal the sample from outside interference. A liquid sample can be applied to this agar film and left to incubate over a period of 1 - 2 days. After this time, the bacteria that were present within the sample will have had enough time to grow into colonies which are visible to the human eye. These colonies can be counted and the original number of bacterium per unit volume can be known from the original sample. The process to count the bacteria samples and categorize them into datasets is often tedious and can take a human operator around 1 - 2 minutes per sample. With sample sizes of 50 - 60, this process can easily take over an hour. The goal of our project is to use an automated system to count these bacteria colonies in sample sets of upwards of 100+ in as little as 20 minutes. This data will be automatically stored and tabulated within the client's original database.

Being a robotics project there are inherent complexities that must be resolved between subgroups such as the electrical control of the mechanical design, computer commands for electrical control and how effective the mechanical design facilitates the overall process. With that, the project must maintain a high level of communication and simplicity so that all groups are able to understand how their design affects the overall system and others work.

Mechanical Sub-System

The mechanical sub-system was designed using industrial examples of delta robots in mind and special attention paid to the accuracy, speed, and cost of the system. Arguably, the most important aspect of our project is that is must be accurate so as not to damage the provided samples and to improve it's overall time cost for this process (less mistakes mean less time needed to recover). The delta robot design was made using many commercial off the shelf (COTS) parts with a few custom designed parts that were manufactured using a 3D printer (the Creality Ender 3 .4mm to be precise).

Electrical Sub-System

The system interacts with the samples to be tested via an end-effector that is moved by three stepper motors in a delta configuration. The goal of the Electrical Sub-System is to enable these interactions to happen at the command of the Computer Science Sub-System. The stepper motors are being controlled via an ATmega 328 microcontroller. The microcontroller is running specialized software which translates coordinates sent from a Raspberry Pi into waveforms representing the direction, distance, and speed a motor should move in. These waveforms are then passed to a stepper driver, which switches power to the motor according to the information within these waveforms. The end result is a stepper motor which can be controlled via a Raspberry Pi. An additional ATmega 328 microcontroller is utilized to control the other electrical aspects of the system as well. Such duties include enabling and disabling the vacuum pump, as well as enabling and disabling the flash for the system camera.

Computer Science Sub-System

The computer science sub-system has three primary functions: automation control, computer vision image analysis and user interface. All three of these functions are implemented on the RaspberryPi microcontroller. The user interface was created using PyQT and is used to input various pieces of data such as how many samples are to be analyzed and also where the user starts/stops/resets the analysis process. The user interface is updated throughout the process and displays the current sample image as well as relevant data about the sample. The automation controls are calculated by the RaspberryPi which then sends the control messages via serial interface to the ATmega328 where they are decoded and used to set the motors to the proper angles. The image analysis is handled using OpenCV and can analyze approximately 5 samples per minute. It counts the number of microbial growths if any are present. After the data has been collected it is output to a .csv file that is easily transferred into a spreadsheet report.

About

EECS 441 - 443 Project folder with files for senior capstone project for Jorian Bruslind, Mack Hall, and Zach Bendt

Resources

License

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Contributors 3

  •  
  •  
  •