PRESENTERS

Department of Computer Science & Engineering

Autonomous Drone-Based Object Detection for Vegetation Hazard Identification and Wildfire Prevention Near Power Lines

Wildfires in the United States have increased in both frequency and severity, with a significant portion attributed to power line failures igniting nearby vegetation. This project develops an autonomous unmanned aerial vehicle (UAV)–based system to detect vegetation encroachment and other fire hazards along rural power line corridors. The platform is built on both a HolyBro X500 V2 ARF carbon fiber quadcopter, equipped with a high-resolution RGB camera, Raspberry Pi processing unit, BerryGPS-IMU module, and telemetry radio for real-time geospatial data logging and an Aero Drone. Flight missions are pre-programmed via QGroundControl to execute autonomous waypoint navigation, capturing continuous video of transmission infrastructure. Video streams are segmented into still frames and analyzed using a custom-trained object detection model to identify hazards such as dead trees, invasive vegetation, branch overhangs, damaged insulators, and conductor contact points. Training datasets are curated from both real-world imagery and generative AI–augmented data, labeled for precise vegetation–power line proximity classification. When hazards are detected, the system logs GPS coordinates and structure identifiers, transmitting them to a central database accessible to utility operators and local authorities for rapid mitigation. The solution operates fully offline, is low-cost, and leverages open-source hardware and software.

RESEARCHERS: Nirupama Balaji, American High School '28; Elizabeth Ashley, Milpitas High School '26

ADVISOR: McMahan Lab, Quantum Computing & Computer Science 

KEYWORDS:  Mechanical Engineering | Robotics | Machine Learning | Wildfire Prevention | Unmanned Autonomous Vehicles 

Department of Chemistry, Biochemistry & Physics  

Cellular Antiproliferative Activity of C-4 Modified Podophyllotoxin Analogs Is Uncoupled from Tubulin Binding Affinity

Antimitotic agents, such as podophyllotoxin and its derivatives, target the most fundamental mechanisms of cell division and have demonstrated broad clinical potential as anticancer therapeutics. Given the pharmacological importance of the C-4 hydroxyl on podophyllotoxin, we sought to establish a structure-activity relationship between chemical modification and in vitro potency. We previously reported that increasing C-4 sterics of podophyllotoxin esters minimally affects cell-free tubulin polymerization but decreases in vitro potency against human colon cancer cells. In this study, the antiproliferative activity of 26 novel carbonate, carbamate, and silyl ether podophyllotoxin analogs was evaluated for inhibition of in vitro cell viability, cell cycle arrest, cell-free tubulin polymerization, immunofluorescence imaging, and docking models. In human and murine colorectal cancer (HCT-116, HT-29, CT-26), lung cancer (A549, Calu-1), triple-negative breast cancer (MDA-MB-231 P), liver carcinoma (HepG2), ovarian adenocarcinoma (SKOV-3), and embryonic kidney cells (HEK-293), the carbonate analogs exhibited greater antiproliferative activity when compared to their carbamate counterparts, and the silyl ether analogs performed the poorest. Most notably, the t-butyl carbonate exhibited the most potent activity with IC50 at or below single digit nanomolar ranges, while the carbamates exhibited IC50 values greater than 42 nM. Correspondingly, the most potent analogs also exhibited potent cell cycle G2/M arrest by flow cytometry, with t-butyl carbonate causing arrest to the greatest percentage (69%), affirming tubulin inhibition as the primary mechanism of action. However, those differences were not reflected in the cell-free biochemical assay, where the best carbonates and carbamates showed similar percentages of tubulin inhibition (86% and 85% respectively), and no correlation was observed between C-4 substituent and tubulin inhibition. Additionally, in silico modeling revealed a minimal difference in each analog’s binding affinity, providing further evidence that C-4 functionalization does not directly alter a compound’s ability to bind to tubulin. To investigate whether nonspecific binding to serum albumin may contribute to the observed discrepancy between cell-based and cell-free potency, experiments are currently underway comparing compound activity in the presence and absence of albumin. Collectively, our results suggest that absolute binding to tubulin is not the primary determinant of biological activity for C-4 analogs of podophyllotoxin across a broad panel of cancer cell lines, and that the impact of structural changes at C-4 on biological potency is agnostic to logP and steric bulk. Altogether, the structure-activity relationship described enables future development of podophyllotoxin-based antimitotic agents for cancer.

RESEARCHERS: Anna Gribok, Mission Early College High School ‘27; Rebecca Bellis, Alameda High School ‘26

ADVISOR: Njoo Lab Synthesis | Physical Organic Chemistry | Catalysis | Chemical Biology | Spectroscopy | Medicinal Chemistry 

Yamamoto Lab Biomaterials Engineering, Materials Science

KEYWORDS: Natural Products | Medicinal Chemistry | Chemical Biology

Department of Computer Science & Engineering

Measuring Air Pollution with the use of Unmanned Aerial Vehicles

Greenhouse gases such as CO₂, CH₄, and NOₓ are powerful drivers of climate change and loss of biodiversity despite monitoring networks now available being costly, sparse, and incapable of resolving fine‐scale complexities in emissions. Using small gas sensors on unmanned aerial vehicles (UAVs) with machine‐learning models is thwarted by difficulties in sensor calibration, processing data for flight operations, and adaptive navigation but allows rapid, high‐resolution mappings of pollutant concentrations. We present a UAV system for monitoring pollutants here in which ground‐truths are used to train machine‐learning models for converting raw sensor data to right gas concentrations, adaptive navigation for targeted hotspots in flight operations, and high‐resolution mappings of emissions with a scalable cost‑effective solution over labor‑based monitoring with recommendations for priority work towards mitigation.

RESEARCHERS: Christon Rex, California High School '28; Yash Shekhawat, Mission San Jose High School '28;  Alyssa Kwon, Dougherty Valley High School ‘27

ADVISOR: McMahan Lab, Quantum Computing & Computer Science 

KEYWORDS:  Unmanned Aerial Vehicle | Gas Sensors |Air Pollution | Emission Mapping | Autonomous Navigation | Greenhouse Gases | Environmental Science

Department of Chemistry, Biochemistry & Physics  

Bridging the broad spectrum of chemistry: Amino triester lipids as biodegradable surfactants for drug delivery and precise control of quantum dot formation

Aminotriester lipids have been used as surfactants in the formulation of nanoparticles, liposomes, and other macromolecular structures for delivery of therapeutic nucleic acids such as mRNA and siRNA, aptamers, and oligonucleotides across the cell membrane. While we found that these molecules have limited effect on the delivery of small molecule anticancer drugs, inspired by structural analogy between our aminotriester lipids and tri-n-octyl phosphine (TOP), a tri-alkyl surfactant ligand used in the synthesis of colloidal quantum dots (QDs) for modulation of nucleation and growth, we investigated the partial substitution of TOP with our aminotriester lipids as surface ligands in the modulation of quantum dot growth rate and stability over time. Through a quantitative fluorescence spectroscopy study, we discovered the aminotriester lipids functioned as a stabilizing agent that slowed the growth rate during synthesis, leading to a significant blue shift in the resulting spectra, and decreased the rate of QD aggregation over time in ambient conditions. 

RESEARCHERS: Elizabeth Yu, Mission San Jose High School '26

ADVISOR: Njoo Lab Synthesis | Physical Organic Chemistry | Catalysis | Chemical Biology | Spectroscopy | Medicinal Chemistry 

KEYWORDS: Lipid Synthesis | Quantum Dots | Fluorescence Spectrometry | Inorganic Chemistry

Department of Computer Science & Engineering

Quantum Error Detection and Mitigation

Quantum computing has the potential to significantly improve computational tasks. Unfortunately, errors due to outside interactions cause the data to be noisy. Our group is focusing on finding a way to decrease these errors by reverting the data back to what it was before the noise. To simulate realistic errors, we used the qsurface surface code simulator in Python. We use both Convolutional Neural Networks (CNNs) for preprocessing and feature extraction, and Graph Neural Networks (GNNs) for predicting and correcting errors in the surface codes. By training and evaluating the CNNs and the GNNs, we can increase performance in quantum computing. These changes will allow for major developments in the field and will be a significant contribution to making an accurate quantum structure.

RESEARCHERS: Satvik Dronavalli, Independence High School '26; Vihaan Krishnakumar, Archbishop Mitty High School '26

ADVISOR: McMahan Lab, Quantum Computing & Computer Science 

KEYWORDS:  Quantum Computing, Surface Codes, Deep Learning, Python

Department of Chemistry, Biochemistry & Physics  

Mechanistic Investigation Into Phenolic Ester Mediated Site Selective Amine Acylation 

The formation of amide bonds is frequently employed in the synthesis of biologically active compounds, but due to the similar reactivity of different amines, the ability to obtain high yields in reactions for polyamine-containing substrates is challenging. While the use of protecting groups has frequently been employed to solve this challenge, developing reagents for exquisite selectivity for the selection of one amine nucleophile over other nucleophiles has the potential to greatly decrease the step count and provide expedient access to mono-acylated products of polyamine substrates. Here, we prepare a library of 10 para and meta-substituted phenolic esters with a variety of electron-withdrawing and electron-donating groups and demonstrate their unique selectivities for amine nucleophiles that were previously not possible to achieve through conventional amide bond-forming reactions. The mechanistic basis for this selectivity is further justified through the use of real-time reaction kinetics by 19F Benchtop NMR. 

RESEARCHERS: Ellie Leo, Aragon High School '28

ADVISOR: Njoo Lab Synthesis | Physical Organic Chemistry | Catalysis | Chemical Biology | Spectroscopy | Medicinal Chemistry 

KEYWORDS: Organic Synthesis | Chemoselectvity | Phenolic Esters | 19F Benchtop NMR

Department of Computer Science & Engineering

Python-Based Architectural Framework of End-to-End BB84 Noise Simulation

Quantum Key Distribution (QKD)  provides a novel alternative to Classical Key  Distribution in protecting secure  communications, but poses challenges in terms of  research and widespread adoption. While the  resource-intensive nature of QKD setups limit  experimental studies to specialized laboratories,  we present an end-to-end simulation that models  photon generation, transmission, and detection,  thereby lowering the barrier to testing hypotheses  regarding experimental setups under realistic  conditions. In this study, a fiber optic-based  simulation is utilized along with a quantum  circuit and a detailed noise model within the BB84  protocol to recreate an experimental setup. Our  BB84 simulation retained ≈50% of transmitted  bits after sifting, matching the experimental  expectations. Our model incorporates  customizable noise, such as multi-photon  emissions and channel disturbances, directly into  the key exchange process, allowing for an  approach that utilizes simulated fiber optics and  theoretical noise models to replicate physical  conditions. Our findings support the development  of QKD as a scalable solution for securing the  quantum internet and future communication  networks. 

RESEARCHERS: Samiha Das, Archbishop Mitty High School '28; Elaine Huang, Harker Upper School '28

ADVISOR: McMahan Lab, Quantum Computing & Computer Science 

KEYWORDS:  Quantum Key Distribution | Qiskit | Quantum Computing | Communications | BB84

Department of Chemistry, Biochemistry & Physics  

Automated Quantification of GUV Popping Assays via YOLO and DINO Image Analysis

Giant unilamellar vesicles (GUVs) are widely used model membrane systems for studying the stability and rupture of membranes, as well as responses to chemical or physical perturbations. Popping assays serve as the most common method for quantifying vesicle failure across attacking compounds. However, these assays are usually analyzed through the use of manual counting or threshold-based methods, both of which are either inefficient or inaccurate. In this study, we develop a deep learning image analysis system to automate the quantification of GUV popping assays, utilizing YOLOv11-based object detection, intensity calibration, geometric filtering, and statistical processing to distinguish GUVs from other objects; most notably MLVs and MVVs. Simultaneously, a regression-based DINOv2 model is trained to estimate GUV counts from diverse training images.  This enables the comparison of detection and global prediction architectures. Together, this work demonstrates a much more efficient, unbiased analysis of GUV popping assays and highlights the potential future use of computer vision to streamline the process further.

RESEARCHERS: Aarav Anand, Lynbrook High School '27

ADVISOR: Njoo Lab Synthesis | Physical Organic Chemistry | Catalysis | Chemical Biology | Spectroscopy | Medicinal Chemistry 

KEYWORDS: Computer Vision | Deep Learning | DINO | GUV | Popping Assay | Regression | YOLO

Department of Computer Science & Engineering

Using Quantum Neural Networks (QNNs), Quantum Vision Transformers (QVT), and the Mathematical Morphological Reconstruction Algorithm (MMR) for Brain Tumor Detection

Brain tumors affect millions around the world, so detection is critical to helping doctors determine treatment. Currently, radiologists manually identify tumors through MRI (Magnetic Resonance Imaging) scans; however, this poses several limitations: it creates a heavy reliance on the experience of radiologists, has become increasingly costly and time-consuming, and is not as accessible to areas that lack the necessary resources and doctors. With the advancement of deep learning algorithms, a more accessible and efficient solution is possible. Given the existing research in classical Convolutional Neural Networks (CNNs) for tumor detection, Quantum Convolutional Neural Networks (QCNNs) and Quantum Vision Transformers (QVT) offer a promising approach to the problem. Mathematical Morphological Reconstruction (MMR), another image processing method, provides a relative metric for success in the QCNN, and is another classical alternative to CNNs. This research compares the accuracy and computational speed of the MMR, QCNN, QVT, and CNN algorithms to determine whether introducing a quantum aspect presents any noticeable advantage. To build these models, extensive datasets of MRI brain scans were collected. The MMR algorithm involved applying various techniques such as dilation, erosion, and skull stripping through OpenCV2's morphology functions. The QCNN algorithm utilizes quantum power to encode the data into a parametrized quantum circuit and apply convolutional and pooling layers. In terms of future steps, QVTs will be implemented with QCNNs for higher spatial understanding. So far, our results indicate that the MMR algorithm achieved up to 92% accuracy. These results will be compared with the accuracy of the QCNN, QVT, and CNN algorithms.

RESEARCHERS: Rutvi Mudalagi, Amador Valley High School '27; Vivaan Sheoran ‘28 Leigh High School

ADVISOR: McMahan Lab, Quantum Computing & Computer Science 

KEYWORDS:  QCNN | QVT | Brain tumor

Department of Chemistry, Biochemistry & Physics  

Reactivity-informed Pharmacophore Editing and Biological Evaluation of Andrographolide and its A-ring Analogs

Natural products and their analogs have long served as inspiration for the exploration and development of small molecules with therapeutic significance. One such compound is andrographolide, a labdane diterpenoid extracted from the plant Andrographis paniculata, which has been extensively studied as an anticancer therapeutic. It is known to function putatively through covalent inhibition of NF-kB, a transcription factor at the crossroad of a myriad of cell signaling pathways that modulate tumor survival and metastasis. Among several targets for A-ring functionalization, we were intrigued by early isolation of an A-ring oxetane analog that is a biosynthetic byproduct isolated from A. paniculata (Jantan et al., Phytochemistry 1994), whose synthetic preparation and biological properties are not well described. After investigation of several synthetic route candidates, we identified an efficient route to access the A-ring oxetane with three chromatographic purifications. En route of investigating A-ring functionalization, we synthesized a variety of A-ring analogs and we describe mechanistic insight into A-ring reactivity of andrographolide and its analogs.

RESEARCHERS: Abigail Yee, Milpitas High School '27; Ruirui Liu, Mission San Jose '27

ADVISOR: Njoo Lab Synthesis | Physical Organic Chemistry | Catalysis | Chemical Biology | Spectroscopy | Medicinal Chemistry 

KEYWORDS: Andrographolide | Natural Products 

Department of Computer Science & Engineering

Comparative study on three machine learning models in novel autonomous drone-based analysis of bike lane infrastructure

Ensuring the safety of bikers on bike lanes is essential to many cities, but monitoring the bike lanes manually is hard and time consuming. This research proposes a way to use drones to effectively inspect bike lanes using sensors  and cameras. The drones will use sensors to autonomously capture images to identify cracks, debris, and other obstructions. By using drones, this will result in less casualties on the road and a more efficient and safe experience for the riders. Drones are more cost effective, faster, and overall better than manually inspecting the bike lanes. The expected outcome of this project is that bike lanes will be much safer to ride on and accidents will greatly reduce and it will improve the cyclists safety.

RESEARCHERS: Tirth Suba, Irvington High School '29; Aarnav Gutti,  Irvington High School '29

ADVISOR: McMahan Lab, Quantum Computing & Computer Science 

KEYWORDS:  Machine learning (ML), Drone-based inspection, Bike lane safety, Autonomous monitoring

Department of Chemistry, Biochemistry & Physics  

Scalable formal synthesis of (+)-etomoxir and NMR-enabled optimization of a catalytic, aerobic oxidation and Aldol-Luche sequence

Etomoxir is a covalent inhibitor of CPT1, a transmembrane mitochondrial protein that acts as the rate-limiting enzyme for fatty acid oxidation. This enzyme plays a major role in metabolic diseases such as diabetes, where regulation of fatty acid biosynthesis and β-oxidation kinetics through CPT1 are effective treatments for such diseases. The 4-Cl phenolic ether on (R)-(+)-etomoxir is a key SAR hotspot for enabling isoform selective inhibition of CPT1. Previously reported syntheses either require early installation of a 4-Cl phenolic ether which precludes the potential for late stage aryl substitution, or employ large scale pyrophoric reactions in early synthetic operations which are challenging to scale. We demonstrate the scalability of a new synthetic route to intercept a late-stage allylic alcohol in route to (R)-(+)-etomoxir. Notably, our alternate retrosynthetic disconnection, which proceeds through a catalytic aerobic oxidation and a one-flask tandem aldol condensation- reduction sequence, to install a key allylic methylene, avoids pyrophoric materials such as n-butyllithium. With a scalable synthesis of a key diversifiable intermediate in hand, our laboratory is currently preparing a library of diverse (R)-(+)-etomoxir analogs to more fully interrogate the SAR of the aryl ring in CPT1 inhibitory activity.

RESEARCHERS: Anca Stefan, Cambrian Academy '26

ADVISOR: Njoo Lab Synthesis | Physical Organic Chemistry | Catalysis | Chemical Biology | Spectroscopy | Medicinal Chemistry 

KEYWORDS: Formal synthesis | Catalysis | Spectroscopy

Department of Computer Science & Engineering

Fracture morphology and mechanical properties of 3D-printed PLA: orientation and cross-sectional size study

With the growing popularity of additive manufacturing, an alternative technology to traditional manufacturing, it’s application has been used in a range of fields from structural and civil engineering to dental and precision manufacturing. Recent studies have shown that the mechanical properties of 3D printed parts depend on the printing parameters, color and brand of the feedstock, and brand of the printer. We believe that print orientation is another contributing factor to the properties of 3D printed parts. To assess this hypothesis, we compared the mechanical properties evaluated by ASTM tensile test between two printing orientations(vertical and horizontal) of varying thicknesses(7mm and 3.2mm). From our results, we saw a fine split between the mechanical properties of each orientation and each size sample. To analyze these results we will conduct more literature search and discussion. 

RESEARCHERS: Annika Hegde, Leland High School '27; Jiya Sahlot, Evergreen Valley High School '27; Kavya Karthik, Milpitas High School '28; Runying Gao, Palo Alto High School '27

ADVISOR: Starostina Lab, Materials Science 

KEYWORDS: 3D-PLA | Surface Roughness | Fractured Surface | Tensile Testing | Mechanical Properties| Materials Science

Department of Computer Science & Engineering

Machine learning in SEM image analysis of nanostructures : count, morphology, and size

Nanostructures are usually less than 100 nanometers in size, and because of their small size and high surface-to-volume ratio, they exhibit unique properties that do not appear in bulk materials. Understanding their unique properties is essential for industries focused on semiconductors, medicine, and biochemistry. Scanning Electron Microscopy (SEM) is a technique that uses a beam of electrons to scan the surface of a material to create highly detailed images of nanostructures. Some problems that occur in Scanning Electron Microscopy is that its process is both difficult and time-consuming. In our project, we plan to automate the SEM process with the utilization of AI to minimize human intervention, since humans are more prone to errors than machines. So far, we have developed SEM image analysis models focused on the detection of random nanospheres, semi-aligned nanospheres, nanotubes, and nanowires.

RESEARCHERS: Andrew Kim, Corona Del Mar High School '26; Matthew Nadavallil, California High School 29'; Sanvi Desai, Cupertino High School 27'; Carter Tsao, Harvard-Westlake Upper School 27'

ADVISOR: Starostina Lab, Materials Science 

KEYWORDS: Machine Learning | Materials Science | Nanostructures | Scanning Electron Microscopy

Department of Computer Science & Engineering

ADME Evaluation of Andromethoxy, an Andrographolide Analog 

Andrographolide is a therapeutic compound found in Andrographis paniculata, a plant known in traditional Chinese medicine. It possesses anti-inflammatory and anti-cancer properties and can be used to treat the inflammatory effects of numerous diseases [1]. However, the original compound has low bioavailability [1]. Andromethoxy, a chemically modified derivative of andrographolide, has been designed to improve these properties. This project aims to evaluate Andromethoxy’s absorption, metabolism, distribution, and excretion (ADME) characteristics through 4 assays. By comparing its performance to the parent compound, we hope to identify whether methoxylation improves membrane permeability, metabolic stability, and overall pharmacokinetic behavior. 

RESEARCHERS: Amrutha Boggavarapu, Notre Dame San Jose High School '27; Raya Sai, Notre Dame San Jose '27; Ojasvi Dharnidharka, Leland High School '28;  Pearl Shah, Mountain House High School '27; Hasan Modan, Joseph A Gregori High School' 27

ADVISOR: Salgotra Lab, Analytical Biochemistry & Phamarcokinetics Lab

KEYWORDS: Drug Design | Pharmacokinetics | ADME Analysis | Drug Metabolism

Department of Chemistry, Biochemistry & Physics  

Progress towards the asymmetric total synthesis of the Sporovexin natural products

The evolutionary competition between fungal species has yielded an abundance of small-molecule antimicrobial natural products. Natural products serve a key role in medicinal chemistry, with natural product-inspired compounds comprising up to 50% of FDA-approved drugs over the last thirty years, and an additional 5% directly derived from natural sources. Among them, Sporovexins A–C are p-hydroxybenzoic acid metabolites of the fungus Sporormiella vexans that were demonstrated to exhibit antibiotic and antifungal properties in preliminary assays. Despite their potential, these molecules have yet to be synthesized, with prior literature exclusively focusing on their direct isolation from Sporormiella vexans, and no further assays were ever completed to further understand their properties. Further, we aim to confirm the absolute stereochemistry of this family of molecules, which is currently unknown. Here we present the synthesis of Sporovexin A and B, as well as two novel des-methyl analogs of the Sporovexin family from commercially available starting materials to probe the specific effects of these functional groups on antimicrobial activity. Notably, we deploy an asymmetric aldol addition to simultaneously install the stereocenters of Sporovexin B with absolute enantio- and diastereoselectivity.

Complex molecules with multiple stereocenters, including α-phenoxy carbonyl derivatives, serve as valuable intermediates in organic synthesis. Due to their atom and step economical nature, asymmetric aldol additions are a particularly attractive option for forming carbon-carbon bonds while retaining stereoselectivity to make such molecules. One application of this transformation is in the total synthesis of the Sporovexin family of natural products, which our lab has taken interest in because of their previously reported antibacterial and antifungal properties. Despite the biological promise of the Sporovexin family, no total synthesis has been reported to date, and their structures and activities remain underexplored. Our lab utilized a magnesium chloride catalyzed aldol addition to synthesize Sporovexin B from methyl paraben in just four steps. The use of magnesium chloride as a Lewis acid catalyst promotes high diastereoselectivity under mild conditions, in contrast to traditional boron and titanium mediated aldol reactions that require extreme air free conditions and highly toxic reagents. This method is also both cost effective and comparatively green. In addition, we synthesized des-methyl variants of both Sporovexin A and B, and we aim to confirm the stereochemistry of Sporovexin B. This effort allows us to better understand the biological activity of these molecules, which will be further investigated through bacterial assays. 

RESEARCHERS: Jay McChesney, Dougherty Valley High School '26; Elias Yao, The King's Academy '26

ADVISOR: Njoo Lab Synthesis | Physical Organic Chemistry | Catalysis | Chemical Biology | Spectroscopy | Medicinal Chemistry 

KEYWORDS: TBA