Old Projects

Quantum Computer Noise Reduction

Sponsored by Prof. HongWen Jiang

Manager: Grant Mitts 

Contact: grantmitts@g.ucla.edu

Year Completed: 2017-2018

Learning Objectives: Construction, Materials Engineering, Data Analysis

The semiconductor qubit measurements in Prof. HongWen Jiang’s lab are extremely sensitive to their environment. Our apparatus will focus on reducing acoustic interference and include a measuring device to determine the efficacy of our dampening apparatus. In this project, team members will learn to design and construct an apparatus to reduce the noise from his lab’s helium compressor, and analyze data from Prof. Jiang’s experiment with and without their dampener, to improve their final product.


Froot Loop Dispenser

Sponsored by Prof. Mayank Mehta

Manager: Krish Kabra 

Contact: krish97@g.ucla.edu

Year Completed: 2017-2018

Learning Objectives: Microcontrollers, Electronics, Apparatus Design & Construction

Prof. Mayank Mehta’s research focuses on answering the question of how the mind emerges from activities of ensembles of neurons. In order to investigate this, they perform experiments on rats in unusual environments, which includes a symmetric room that has no visual landmarks that a rat can use as a reference point to orient oneself. The purpose of this experiment is to understand how a rat’s brain maps such an environment with no visual aid. As part of the experiment, Froot Loops are dropped into the room at random intervals in random locations to learn how the rat responds. However, there is currently no way of dispensing the loops without having an experimenter within the room, which makes the experiment much more dubious and time-consuming. The goal of our project is to create a dispensing mechanism, and use a microcontroller to operate it, which would eliminate the need to have an experimenter in the room while running the experiment.


Satellite Launching and Motion Simulation

In House

Manager: Suyash Kumar

Contact: suyashsep12@gmail.com

Year completed:  2018-2019

Learning Objectives: Simulation, Programming (Python)

Due to the mathematical complexity of increasingly realistic physical systems, programming has become an indispensable tool in developing a greater understanding of physical phenomena through the creation of simulations. For example, solving Newton’s Laws by hand to obtain the equations of motion can become extremely difficult, but creating simulations to perform numerical calculations to obtain a good approximation requires a few lines of code. The example that we focus on is the process of launching a satellite from the surface of the Earth. By performing numerical simulations of this classic physics problem, our members develop the skills to simulate more complex systems that would otherwise be difficult to envision.


PID w/ Microcontroller

An Original In-House Upsilon Lab Project

Manager: Helena Huang

Learning Objectives: Microcontrollers, Electronics (Design, Soldering)

Year completed:  2018-2019

A Proportional-Integral-Differential (PID) controller is a device used to make educated guesses with a system’s history to predict its future and to control physical parameters using that knowledge. For example, most cars use PID controllers to control their speed – they look at the current speed, where it’s going (derivative) and where it’s been (integral). In this project, team members will design a car together, and learn how to program a microcontroller (MCU), specifically an Arduino development board which utilizes the ATmega328p, to control the movement of the car so that it maintains a certain distance with the object in front of it. This will include how to interpret signals from a distance sensor, how to store and manipulate data for PID control, and how to output a signal to control the motion of the car.


Solar Analysis

In House & Partnership with the REA Solar Team

Manager: Emma Peavler

Contact: lizzyp@g.ucla.edu

Year Completed:  2018-2019

Learning Objectives: Simulation, Programming (Python, Matlab, and/or Mathematica), Report Writing

As the effects of non-renewable energy resources (i.e. coal, oil etc) are compounding the issue of global warming, the creation of practical and manufacturable renewable energy sources is necessary for the future of humanity. This experiment group will be designing, simulating and creating solar energy storage devices using lithium ion battery cells (18650) and various solar panels. In this experiment team members will learn and use basic programming knowledge to simulate the efficiency of various solar panels and optimize a design that allows for the maximum return on investment for each solar pack. Members will also be able to engineer and build the energy storage packs for optimized use.


Cyclotron Motion Simulation

In-HousE

Manager: Jared Rivera 

Contact: jaredrivera2314@gmail.com

Year completed:  2018-2019

Learning Objectives: Programming (Python)

Cyclotron motion is the motion that charged particles undergo in a magnetic field such that they move outward in a spiral path. Team members use Python to simulate the motion of charged particles undergoing cyclotron motion. They will then apply this simulation to physical systems to predict outcomes of potential experiments, one example of which includes determining plasma wave behavior in the atmosphere using cyclotron accelerators, and another example being determining particle trajectories in a particle collider.


DECODING BRAIN SIGNALS WITH DEEP LEARNING

Manager:  William Zhu – In-house

Contactwilliamzhu@g.ucla.edu

Learning Objectives:  Programming (Python, Keras, and Tensorflow), Neural Networks, Deep Learning

APPLICATIONS: OPEN

The inner workings of a brain is one of the greatest mysteries yet to be completely solved by science. Researchers need to make sense of an immense amount of brain signals in order to understand the mechanisms behind higher-order functions—such as perception, feelings, memory, or the sense of self—produced by this three-pound network of neurons. The neural networks used in deep learning, themselves inspired by the brain and designed to process large amounts of information, fit well with this task.

In the fall quarter, we will first gain experience with deep learning through practicing on simple image data sets with convolutional neural networks (CNNs). We will gain exposure to other types of neural net architecture as well, such as recurrent neural networks (RNNs) for time-domain signal analysis. Then, starting in the winter quarter, we will apply the skills and knowledge we have acquired to build a neural net that decodes behavioral and cognitive states from fMRI images (all data will be from public domain datasets, so we won’t work with actual fMRI machines. Sorry to disappoint you!). The exact research topic is subject to modification, depending on how everyone feels about coding or the ideas behind deep learning. At the very least, students will learn very important skills that are both handy in laboratories and sought-after in the job market, including Python, Keras, and TensorFlow.


DECODING BRAIN SIGNALS WITH DEEP LEARNING

Manager:  William Zhu – In-house

Contactwilliamzhu@g.ucla.edu

Learning Objectives:  Programming (Python, Keras, and Tensorflow), Neural Networks, Deep Learning

APPLICATIONS: OPEN

The inner workings of a brain is one of the greatest mysteries yet to be completely solved by science. Researchers need to make sense of an immense amount of brain signals in order to understand the mechanisms behind higher-order functions—such as perception, feelings, memory, or the sense of self—produced by this three-pound network of neurons. The neural networks used in deep learning, themselves inspired by the brain and designed to process large amounts of information, fit well with this task.

In the fall quarter, we will first gain experience with deep learning through practicing on simple image data sets with convolutional neural networks (CNNs). We will gain exposure to other types of neural net architecture as well, such as recurrent neural networks (RNNs) for time-domain signal analysis. Then, starting in the winter quarter, we will apply the skills and knowledge we have acquired to build a neural net that decodes behavioral and cognitive states from fMRI images (all data will be from public domain datasets, so we won’t work with actual fMRI machines. Sorry to disappoint you!). The exact research topic is subject to modification, depending on how everyone feels about coding or the ideas behind deep learning. At the very least, students will learn very important skills that are both handy in laboratories and sought-after in the job market, including Python, Keras, and TensorFlow.