Water Engineering For Good - 2026

Behind every innovative engineering project lies a tale of experimentation, failure, and success. Our group spent several months under tight deadlines to address one of the biggest problems in California: the water footprint caused by the AI revolution.

The following timeline documents what we have done throughout the recent two months in preparation of WE4G:

February: The Deep Dive & The Detection

To begin with, we had to investigate the "hidden" price of AI development. Our analysis showed us that even though each individual student uses very little amount of AI, the vast data centers supporting it in Southern California use millions of gallons of freshwater for cooling each day.

February: The Deep Dive & The Detection

To begin with, we had to investigate the "hidden" price of AI development. Our analysis showed us that even though each individual student uses very little amount of AI, the vast data centers supporting it in Southern California use millions of gallons of freshwater for cooling each day.

Late February: Engineering the “Closed Loop”

Having identified the challenge, we moved into the concept stage. After brainstorming ideas ranging from a seawater heat exchanger to immersion cooling, we arrived at a holistic solution: the Closed-Loop Purification Skid.

To illustrate our solution, we utilized Text-to-CAD to create a visualization of a modular attachment that could be affixed to any cooling tower.

March: Building a Benchtop Model, and Test It for an Uncertain World

March is the month in which heavy-liftings were done. 

Theories have their limits. To verify our theory, we constructed a working prototype for less than $20 through the use of fine silica sand, carbon, and cotton. We used salt, pepper, and food coloring to simulate "dirty" blowdown water, achieving a starting point of 584 ppm.

Through our benchtop filter, we managed to remove only 3.8% of Total Dissolved Solids (TDS), with an average of 561.8 ppm. Although the water became visually cleaner, the salt still existed in the water. Our failure became our greatest triumph since it provided us with concrete proof that filtration alone is insufficient.

Engineering is not only about the ideal situation but also uncertainty. For most of the month, we were performing Monte Carlo simulations of the performance of our device against variable thermal loads and humidity levels at the server room.

The Outcome: In more than 20 "scenarios," our model showed that on average our device could save up to 0.390 units per day.

April: "The Last Drop" Video Project, and Our Presentation

Finally, we turned towards the role of being advocates. We made use of generative AI to make a social media/advertisement video titled "The Last Drop" (powered by Veo – Google Labs). The video includes a story about a traveler from the year 3026 who comes back to give us a message about a world without water.

We ended our process with presenting our findings to the MWD Engineering Panel. Hopefully you didn't miss it; it's right at the top of this page!

Acknowledgement

Without the backing of our remarkable community, this project will merely be a "hallucination". First, we would like to express our thanks to all our dedicated team members, Xiaomai Cao, Linsey Lin, Enya Jiang, Derek Wang, Alyssa Guo, Dony Li, and Michael Yang, who are devoted to following the "Verification First" criteria and contributing their unique talent pool. In particular, we extend our gratitude to our club director Dr. Fubing Peng, who successfully guided our team's ship through a condensed timeline from February to March while transforming our blue sky brainstorming ideas into an actionable industrial design.

Furthermore, we are thrilled at having joined in the Water Engineering 4 Good 2026 initiative this year. We would like to convey our sincere appreciation to the Metropolitan Water District (MWD) Board of Directors, who gave us the opportunity to explore and develop an investigation on Southern California's complicated future water problem within this year's unique AI-themed challenge. Last but not least, we would like to thank MWD engineering committee members for their invaluable feedback and guidance toward the gap of benchtop prototype and real-world adaptation due to climate change.

We are proud to be part of a generation closing the loop on sustainable technology.