Project

# Title Team Members TA Documents Sponsor
24 Distributed Light System using Voice Recognition
 Anish Naik
Anshul Goswami
Walter Tang
 Jeff Chang design_document3.pdf
final_paper2.pdf
photo1.png
photo2.png
presentation1.pptx
proposal2.pdf
video
# TEAM MEMBERS:
Anshul Goswami (anshulg3@illinois.edu)

Walter Tang (waltert3@illinois.edu)

Anish Naik (anishn3@illinois.edu)

# PROBLEM:
Having voice recognition can be very powerful to help with mundane tasks. For example, maybe you want to turn on a light in one part of the house without having to walk there or having to find the light switch in the dark.

Although Alexa can do something similar, people have privacy concerns about it. Our system offers local processing of voice data so that none of your voice recordings leave the area. This solution offers dedicated functionality which does not require a more complex home system if you don’t want to buy into a larger ecosystem of products. This solution also allows you to turn on/off specific lamps around the house through voice commands.


# SOLUTION OVERVIEW:
We want to create a base station that will be used as a go between between a system with a speech recognition system and lamps throughout the house. It will be made so that specific lights can be turned on throughout the house using voice commands. We will use Wifi to connect the base station to each of the lamp’s outlets.
# SOLUTION COMPONENTS:
Base Station Box consisting of Microphone, Raspberry Pi, and Microcontroller: This will allow us to connect to the various lamps in the house and process whatever voice commands are sent in over the Microcontrollers that are connected via Wifi

Lamps Plug in Box:
Connects the Outlet to the Lamp Plug, and will allow us to turn the specific lamps on or off using a MOSFET. They’ll also have a microphone that can connect to the Base Station with the microcontroller that’s connected to Wifi.

Microphones:
This will be used to pick up the voice of the speaker, and connect it to the systems so it can activate it.

Raspberry pi:
We will either create our own Voice Recognition Module or find software that can do it. Otherwise we will just have it compare the voice to preset voices that we already put in place. This will then send a signal to the Microcontroller based on what the microphone heard

Microcontroller (ESP32):
Connects the Raspberry Pi and the lamp plug in boxes to Wifi which allows them to communicate.

# CRITERION FOR SUCCESS:
Our project contains several boxes which connect the lamps to the outlet with the MOSFET, and then one main box which will house the raspberry pi. Our project will allow us to talk to the main voice station and be able to turn on other lamps throughout the house.

We can demonstrate this by successfully using the voice commands to turn the lights on and off through the main box.

BarPro Weightlifting Aid Device

Patrick Fejkiel, Grzegorz Gruba, Kevin Mienta

Featured Project

Patrick Fejkiel (pfejki2), Kevin Mienta (kmient2), Grzegorz Gruba (ggruba2)

Title: BarPro

Problem: Many beginner weightlifters struggle with keeping the barbell level during lifts. Even seasoned weightlifters find their barbells swaying to one side sometimes. During heavy lifts, many people also struggle with full movements after a few repetitions.

Solution Overview: BarPro is a device that straps on to a barbell and aids the lifter with keeping the barbell level, maintaining full repetitions and keeping track of reps/sets. It keeps track of the level of the barbell and notifies the lifter with a sound to correct the barbell positioning when not level. The lifter can use the device to calibrate their full movement of the repetition before adding weight so that when heavy weight is applied, the device will use data from the initial repetition to notify the lifter with a sound if they are not lifting or lowering the barbell all the way during their lift. There will be an LCD screen or LEDs showing the lifter the amount of repetitions/sets that they have completed.

Solution Components:

Subsystem #1 - Level Sensor: An accelerometer will be used to measure the level of the barbell. If an unlevel position is measured, a speaker will beep and notify the lifter.

Subsystem #2 - Full Repetition Sensor: An ultrasonic or infrared distance sensor will be used to measure the height of the barbell from the ground/body during repetitions. The sensor will first be calibrated by the lifter during a repetition with no weight, and then that calibration will be used to check if the lifter is having their barbell reach the calibrated maximum and minimum heights.

Subsystem #3 - LED/LCD Rep/Sets Indicator: LEDs or a LCD screen will be used to display the reps/sets from the data measured by the accelerometer.

Criterion for Success: Our device needs to be user friendly and easily attachable to the barbell. It needs to notify the lifter with sounds and LEDs/LCD display when their barbell is not level, when their movements are not fully complete, and the amount of reps/sets they have completed. The device needs to work smoothly, and testing/calibrating will need to be performed to determine the minimum/maximum values for level and movement positioning.