2,893 Inspirational Passion Project Ideas

Turn inspirations into your passion project.

This collection of project ideas, shared by Polygence mentors, is meant to help inspire student thinking about their own project. Students are in the driver seat of their research and are free to use any or none of the ideas shared by their mentors.

People working on laptops
Computer Science

Robotic System Design

Robots have many electromechanical parts that keep them running. Create a robotic system while learning about design tradeoffs that balance hardware, software, and cost restrictions. Your main takeaways will be learning about the sensors and electronics needed in almost every robotic design, from self-driving cars to autonomous drones.

Physics, Engineering, Computer Science, AI/ML

Tony
Tony

Tiny Timely Transit

A project I worked on with a few friends to provide more accurate bus time estimates for the public transit around our school. There are endless systems with lots of data to analyze, and building a project like this can teach you about data analysis (and even machine learning), software development, web development, and several more important skills.

Computer Science, Math, Statistics

Aidan
Aidan

What's the matter?

A literature review of the most motivated models of dark matter, along with current constraints on these models. What's a WIMP? An axion? A primordial black hole? Which of these seem most motivated by current observational data?

Physics, Computer Science

William
William

Build-a-Bike:

Yup, it's exactly what it sounds like! Together, we can build your own personal bike from the ground up. I have built several bikes from old beat up parts that now work good as new. And guess how much it costs: $0. From this project, you'll gain much more than just a new bike; you'll also become an expert in every mechanism on a bicycle (gearing, braking, etc.), so whenever a new problem arises (as it always does with mechanical components) you'll know exactly how to fix it.

Computer Science

Alex
Alex

Using machine learning to predict disease mortality rates from publicly available data and identify potential public health interventions

The Center for Disease Control (CDC) maintains a database of underlying causes of death for each state (CDC Wonder: https://wonder.cdc.gov/ucd-icd10.html). By gathering additional data about individual states (e.g. smoking rate, occupation distribution, education levels, sales of certain products, weather patterns, etc), we can develop a machine learning model to predict death rates for a specific disease (e.g. lung cancer) or multiple diseases. The US government maintains many databases containing variables that can be used as predictors. Following the development of a machine learning model, the student researcher can perform statistical analyses to identify important predictors of the disease and propose potential public health interventions to reduce disease incidence. Students pursuing a project of this type will learn how to access publicly available public health databases, develop machine learning models in Python, and perform statistical analyses to identify significant predictors of disease.

Biotech, Biology, Computer Science, Statistics

Joshua
Joshua

(Computation Bio) Build your own Genetic Analyzer

In this project we will explore how to find which genes and regions are conserved. How to pick a gene or protein of interest and how to analyze genomic sequences. From there this project will give you the tooling to explore other genes and even evolution of proteins. You'll be able to say with what confidence certain proteins evolved from others.

Neuroscience, Computer Science, AI/ML

Steven
Steven

Designing and completing a research project given the resources available to you

Have you ever wanted to be a real-life scientist and complete projects such as my research illustrated above? Don't be discouraged by limited resources! I can work with you to facilitate the use of your community or secondary school laboratories, computer labs and libraries, or those of a nearby (or not so nearby) university or company. In this project, you will design a research project that can be completed at, gather data, and present that data as either a scientific poster or talk. You will learn about the scientific process, how to construct appropriate figures for your data, and how to communicate your results in a formal presentation, such as my research featured above!

Physics, Computer Science, Chemistry, Math

Amy
Amy

Modeling the spread of an infectious disease (e.g., COVID-19)

In this project, we will learn about mathematical modeling and its diverse applications in epidemiology. Pertinent questions may be: Which infectious disease should we model (i.e., the “big picture” context)? What is the nature of the data we are provided with (i.e., exploratory data analyses)? What are the various types of modeling frameworks available to us (i.e., model paradigm)? On what levels can models be compared (i.e., model comparison)? Which steps should be taken to ensure the accuracy of our model (i.e., model checking)? The chosen paths of exploration will depend on the student’s interests. Final Notes: If you have a particular idea for a project in mind, I am happy to guide you throughout your journey. Intellectual curiosity is a vital component of the research process!

Computer Science, Math, Statistics

Abhishek
Abhishek

Positive and Unlabeled Materials Machine Learning - Using Semi-Supervised Machine Learning to Identify and Accelerate New Material Synthesis.

Materials Synthesis is the chemical and physical means of using combinations of atoms and molecules to form novel and useful materials. Through materials synthesis, the pathways of the manufacture of new materials are invented. These combinations of atoms, molecules, and compounds have similar characteristics, but the data can be labeled and unlabeled data. Semi-Supervised Machine Learning can be used to identify and classify other unknown materials with the same similar characteristics. These characteristics can be accurately measured in the unknown materials and we could even better evaluate out how these characteristics interact in unknown materials.

Biology, Physics, Computer Science

Darrell
Darrell

Quants and Poets: An Empirical Analysis of Sportswriting vs. The Underlying Numbers

A quantitative analysis that matches sportswriting with the statistics from games covered, attempting to tease out patterns or biases in the writing using text analysis and contrasting them with the numbers of the team and players.

Computer Science, Economics, Math

Anshul
Anshul

Designing your own autocorrect algorithm: an introduction to algorithms and machine learning

In this project, we will focus on two big ideas: one in computer science and one in machine learning. The first idea is called dynamic programming and is one of the traditional ways in computer science to implement an autocorrect algorithm. We can design it from scratch (this is a project I made at Berkeley that has been given to 10,000 students and counting) or just focus on the algorithm, depending on how much you're looking to learn. Then we will switch gears and use machine learning in order to implement a more accurate version of autocorrect that is able to be personalized to each user. By the end of this project, you should be comfortable with a class of algorithms that are taught to junior and seniors in computer science classes at Berkeley, as well as the fundamentals of machine learning!

Computer Science, Math, AI/ML

Ryan
Ryan

Sentiment analysis of COVID-19 vaccine tweets

Apply data mining to query and synthesize hundreds of thousands of tweets and perform sentiment analysis to compare which COVID-19 vaccine is most promising in different geographic regions. Compare which side effects are most predominant among Pfizer vaccine recipients versus Moderna vaccine recipients. This project would likely culminate in a Medium article which takes the reader through the project from exploratory data analysis to code implementation, and finally a well-articulated discussion of research findings and limitations.

Computer Science, Sports Analytics, Quantitative, AI/ML

Sejal
Sejal

Design and Testing of 3D-Printed Propeller

Additive manufacturing has come a long way, specifically in the hobbyist market. Is it possible to manufacture an aircraft propellor? in this project we will explore the aerodynamics behind propellor forces and use them to design a 3d-printed propellor. Next, experiments will be performed to determine how useful this prototype can be.

Physics, Computer Science, Math

George
George

Real-Time Flight Tracking and Data Processing

For my current research, we deploy ADS-B (Automatic Dependent Surveillance-Broadcast) collectors "in the field" to receive radio signals broadcast by planes flying overhead. All commercial and recreational aircraft are required to be equipped with an ADS-B transponder, which means that we have the ability to track live air traffic in a 100-150 mile radius of our current location! This provides a really interesting source of relevant data that can be analyzed to find all sorts of interesting trends. If you're interested in a project at the intersection of computer science, data science, and aviation, you can contact me and we would work to get you set up with your own data collection platform. In parallel, you would come up with a project proposal indication how you would like to use the data you collect to accomplish a relevant and impactful goal. Possible ideas could include: learning how to use visualization software to develop comprehensive plots of aircraft trajectories, making predictions about communities most likely to be impacted by noise from overhead flights, and many many more! Finally, I would help you dive into the coding techniques and skills you would need to accomplish your goal and would support you as you build your data collection/processing system.

Computer Science, AI/ML

Nicholas
Nicholas

Simulating a Memory Array to Understand Performance Characteristics

In this project, the student would simulate a memory array of their choosing, and try to make estimates on the performance based on circuit simulations. Additionally, the student could try to understand different options for improving these metrics by changing parameters such as the array size and device characteristics. This project requires a deep understanding of Physics E&M (AP Physics C + E&M is sufficient) and should be reserved for more advanced students who are comfortable with resistors/capacitors and calculating currents/voltages in circuits. Calculus is highly beneficial.

Physics, Computer Science, AI/ML

Akash
Akash

Robots, Bubbles, and Efficient Markets

There's an old joke in economics that goes like this. A financial economist is walking with his student, and the student sees a $100 dollar bill on the ground. The student is excited, and goes to pick it up, but the professor stops him and says, "there couldn't be $100 on the ground. If there were, someone would have already picked it up!" The professor is expressing the "Efficient Markets Hypothesis", that claims all stock prices are correct, because if they weren't someone could buy and sell them to pick up $100. Of course, many economists don't believe this! There is a whole school called Behavioral Economics that studies how markets fail, why prices are wrong, and the human psychology that causes these inefficient markets to emerge. Typically, behavioral economists work with volunteers, who trade stocks or play games (not the fun kind). The economist try to create price bubbles or market frenzies, and show how irrational people are. But what about AIs? We could program AI to play in these game environments together, and see if the same bubbles and market frenzies occur. Do they eventually turn into the efficient professor after playing the game many times? Can they adapt when the rules of the game are changed? What can we learn about ourselves by studying how robots behave in similar situations?

Computer Science, Economics, Math, AI/ML, Finance

Jesse
Jesse

Coral Reef Health

Formulating your own research ideas can seem like a daunting task from afar but is a lot more manageable once you start brainstorming. A project idea we could work on can involve using existing datasets to tackle an ecological question. For example, using a dataset to analyze coral reef health over time and assess impacts (i.e., bleaching events, disturbances, high temperatures, etc.) by doing basic statistical analysis. This project will involve learning to use excel and potentially R to run statistics. But don’t be scared about having to use math if you're interested in marine science. There are many resources that can help explain statistics, or we could look at the social dimensions of reefs, management, and conservation by analyzing environmental policy instead. There’s so many approaches that you can take to focus on your own strengths or what you hope to learn and improve on.

Computer Science, Chemistry, Environmental Science

Shalimar
Shalimar

An IoT-powered Framework to Assist Mission Control

First responders and military members risk their lives for us. However, I personally witnessed in several occasions that their mission control operations are done using a white-board. I believe this is not a sufficient method, and a more advanced method is required. Therefore, I propose wearable modules powered by the long-range wireless tool named Sigfox that ranges around a couple of kilometers. The wearable modules will have sensors, such as GPS, and vital health information. First responder information will be transmitted to a mobile application that will display each unit's location on the map and their health status. Currently, there are two startups working in this area, and I believe they're making a very positive impact.

Engineering, Computer Science, Business

Berkay
Berkay

Covid-19 Data set Analyzer

In this project you will use a OOP language such as Java to create a Covid-19 data set analyzer that parses the given data set and computes insightful information such as deaths per capita in the given zipcode, vaccination rates per capita, etc. You will learn how to create this project using N-tier architecture (data, processor, and ui tiers) and writing your own CSV parser from scratch. In the end, you will test your hypothesis against your analysis using the information you have gathered.

Computer Science

Thomas
Thomas

Computer Animated Design & Manufactoring (CAD/CAM)

All fields of engineering rely on digital 3D models of parts. These models are incredibly important all the way from design to production. Fortunately, with today's 3D printers it's easy to create 3D model in your favorite CAD program and then see the final product within hours. Find something around your house that needs to be repaired or modified! You'll design a new part while learning about tolerances, assemblies, and best practices for 3D printed parts.

Physics, Engineering, Computer Science, AI/ML

Tony
Tony