Hi, I'm Yash! I'm passionate about using technology to help combat climate change. I have a research background in artificial intelligence applied to climate problems. I am currently a researcher at the Stanford Sustainable Systems Lab and a summer research fellow at Stony Brook's Geosciences Lab. I've published peer-reviewed research in NeurIPS and was awarded a patent for it. I also have a background in applied computer science. I started participating in Hackathons when I was 10 and am a six-time Hackathon winner. I've also launched two apps on the iPhone App Store and my work has been featured in Vox, the SF Business Times, and the Mercury News.
In my free time, I love being in nature and have visited several national parks across the world including in Peru, Costa Rica, Tanzania, China, New Zealand, India, and the United States.
This website is a summary of all my work. I hope you enjoy!
Accurate waste disposal, at the point of disposal, is crucial to fighting
climate change. When materials that could be recycled or composted get diverted into landfills, they cause the emission of potent greenhouse gases such as methane. Current attempts to reduce erroneous waste disposal are expensive, inaccurate, and confusing.
DeepWaste is an easy-to-use mobile app that utilizes powerful artificial intelligence techniques to provide accurate and instantaneous waste classification into trash, recycling, and compost.
Over the last three years, I've taken DeepWaste from an idea to patented and published research to now being deployed into the real world at several school campuses, starting with Williams College, Stanford University, and The Nueva School. Over 3000 individuals are already using DeepWaste to reduce incorrect waste disposal.
Learn more about the work at https://www.deepwaste.ai/
My work for DeepWaste was selected as one of five finalists worlwide for the Children's Climate Prize. This is an international prize based in Stockholm, Sweden, originally started at the UN Climate Change Conference. Each year they select five students worldwide who have made significant contributions to climate change, previous finalists have included Greta Thunberg for her work on climate strikes. It's exciting to be part of this group and continue expanding the reach of DeepWaste. The finalist ceremony is held in the Stockholm Concert Hall in Sweeden where the Nobel Prize was held and the winner will receive SEK 100,000.
I'm a researcher in the Stanford Sustainable Systems Lab advised by Dr. Ram Rajagopal. My current research is focused on the intersection of AI and sustainability, specifically developing machine learning approaches for accurate waste disposal and mapping renewable energy.
Research Project #1
I started my research on DeepWaste independently in 2018, and over the last few years, it has become a full body of work. I published the original DeepWaste work by myself in NeurIPS, a top peer-reviewed conference for artificial intelligence research (impact factor of 33.49).
I then brought the work to the lab and am now directly collaborating with researchers in the Sustainable Systems Lab and Machine Learning Lab at Stanford on a new research study to compare DeepWaste's AI to human accuracy of waste classification. The results are exciting as DeepWaste is outperforming the average human in terms of accuracy by a significant margin (nearly 50%). Based on the results and novelty of the work, we submitted the research to Nature Sustainability (undergoing review).
See the full list of publications, patents, and awards & honors below!
Research Project #2
I'm also on the EnergyAtlas team at Stanford as the only high school student. The vision of the project is to develop a comprehensive database on renewable energy supply, demand, and transmission in all regions of the world through machine learning. Ultimately, the goal of the work is to enable the efficient and sustainable operation of the world's energy grid.
For the last two years, I have been collaborating with Zhecheng Wang (Ph.D. Candidate in Environmental Engineering) on a research project called DeepTransformer as part of the larger EnergyAtlas work. DeepTransformer automatically generates a granular database of distribution lines, electricity poles, and distribution transformers. We are working on publishing EnergyAtlas in a high-impact journal that will encompass DeepTransformer as a central component.
Research Fellow, Stony Brook University with Professor William Holt (2021 - Present)
I was selected as a summer research fellow at Stony Brook University in the Department of Geosciences where I'm advised by Dr. William Holt. I am working on a project called DeepQuake that applies artificial intelligence for real-time earthquake forecasting. The results and conclusions from the research are exciting since our model is able to accurately forecast earthquakes and outperforms other physics-based approaches traditionally used. I presented my research updates in weekly lab meetings alongside the other Simons summer research fellows in the lab. I am continuing my work into the school year and we are working on submitting a manuscript that describes the results of the research.
SchoolHacks was the Bay Area’s first student-led hackathon with a mission to inspire middle and high school students to tackle humanity’s biggest challenges using technology. One of my goals in designing SchoolHacks was to create a very collaborative and inclusive environment for students. I felt that I had found my voice and passion through participating in my first hackathon when I was younger and I wanted every other student to experience that same empowerment.
I assembled a 30-person leadership team of high schoolers and raised $50K in all from world-renowned venture capitalists and companies. I also organized & led daily sessions and fireside chats with inspiring industry luminaries and assembled a panel of distinguished judges. The event grew into a global movement with 350 students in 8 countries and was eventually incorporated into a separate company.
Learn more at https://www.schoolhacks.org/
Founder of ZeroCarbon non-profit (2020 - Present)
ZeroCarbon is a student-led 501(c)(3) non-profit that works in the areas of waste management and energy to help create local communities that transition towards emitting zero carbon.
Initiative #1: Informal Waste Worker Community
I am working on creating Santa Clara County's first informal waste worker community in partnership with the government - informal waste workers are individuals who make an income by collecting bottles and cans from the road and are typically homeless or on the brink of being homeless. We're running a program in partnership with the city and houseless organizations to clean up litter in downtowns while making informal waste workers a formal part of the economy by paying them $20/hour
Image Credit: Trash for Peace
Initiative #2: Microgrid Program in School Districts
My school had not gotten any new green infrastructure (solar, batteries, etc) for 10+ years. I developed a data-driven economic case to advocate for the use of solar and storage-based microgrids for my school that could power the campus for at least two decades. Presented the analysis to the school board and the blueprint developed is being used to evaluate the economic case of microgrids across other school campuses.
2. Narayan, Yash*; Flora, June; Chad Zanocco, Jeremy Irvin; Rajagopal, Ram (2021) “A Study of Accuracy and Usability of AI-based Waste Classification within a Mobile Device” Submitting to Nature Sustainability [preprint]
3. Narayan, Yash “Solar and Energy Storage Based Microgrids: Data-Driven Optimization and Economic Analysis to Examine Energy Savings from a Microgrid within a School Campus” European Geophysical Union [abstract]
4. Narayan, Yash “DeepWaste: Instantaneous and Ubiquitous Waste Classification using Artificial Intelligence for Combating Climate Change” Published in 2020 Columbia Junior Science Journal [paper]
5. Narayan, Yash*; Kim, Jey; Holt, William “DeepQuake: A Scalable, Open-Source Machine Learning Framework for Fine-Grained Earthquake Forecasting” (Working Paper, Target Journal: PLOS Sustainability)
* denotes first-author
Narayan, Yash, "SYSTEMS, METHODS, AND DEVICES FOR CLASSIFYING AND IDENTIFYING WASTE IN REAL-TIME" (Serial Number: 17/146,041).
A device is disclosed which includes a processor, a screen, and either or both of a camera and a microphone. The device may receive input related to a waste item, identify the waste item, classify a disposal type for the waste item, and provide a suggestion for disposing of the waste item.
Patent Granted June 2021.
1) Finalist (1 of 5 worldwide), International Children’s Climate Prize (2021)
2) National JSHS Finalist in Environmental Science (2021)
3) US Department of Environmental Quality Grant Receipt (2020)
4) 1st Place Chevron Innovation Award at Golden Gate STEM Fair (2021)
5) Ricoh USA Sustainability Award at Golden Gate STEM Fair (2021)
6) Joey Kovacevich Social Innovation Fellowship, The Nueva School (2019)
7) Six-Time Hackathon Winner, Apple iOS Developer Camp, the largest and oldest iPhone hackathon (2014-2019)
SF Business Times, “Yash Narayan, the teenager hackathon master” [article]
The Mercury News, “Bay Area’s younger generation doing its part in coronavirus fight” [article]
Vox, “Silicon Valley Teens on the Future of Technology” [article]
Medium, “DevCamp 2017” [article]
Harker News, “Grade 5 Student Wins Award for Designing Anti-Bully App” [article]