Skip to main content

AI & Science

Below are project reports from students and researchers building and using the Sage Grande Testbed. Each node in Sage Grande collects data from cameras, microphones, LiDAR, and environmental sensors, and runs models directly at the edge, so the science can happen where and when the data is created rather than after the fact.

The work below spans many domain sciences, the AI methods that enable them, and the edge infrastructure that supports it all. Also included are projects from Sage Labs, where the team builds and tests new tools for the testbed, and the use cases that show how Sage supports a whole area of science.

These reports are written by students at every level, from early undergraduates through graduate students and postdocs, along with research scientists and occasional collaborators.

AI methods & data
Computer Vision27
Object Detection & Tracking10
Foundation Models & LLMs10
Multimodal Learning7
Self-Supervised Learning3
Acoustics & Audio5
Science domain
Atmospheric Science13
Ecology & Biodiversity10
Water & Hydrology4
Wildfire4
Urban & Transportation5
Edge systems
Edge AI26
Sensors & Instrumentation11
Networking3
Data Management8
Scheduling & Orchestration3
Energy & Resource Management4
Privacy1
Type
Sage Lab6
Use Case2
Sage Summer Camp14

2026

Acoustic Bat Detection

Air Quality Intelligence

Bat Activity Monitoring

Biotic Interaction

Dangerous Animal Detection

Drought Early Warning

Image Search at the Edge

Example graphs showing results from the image search benchmarking system.
Sage Lab

Image Search Benchmarking

Lightning-Ignition Watch

Long-Range Low-Power Networking

Manoomin Vision

Mortimus Urgent Scheduling

Sage-NDP-SciDx MCP

Speech Redaction at the Edge

view of volcano from node in Hawaii
Sage Lab

Volcano, Fire, and Anomaly Detection

Weather Image Classification

2025

Application-Agnostic Dynamic Data Collection

Autonomous Camera Control

screenshot of image searching for volcanoes
Sage Lab

Image Search

screenshot of prompting with mcp
Sage Lab

MCP (Model Context Protocol)

icon representing next-generation Sage Grande edge nodes
Sage Lab

Next-Generation Nodes for AI

deer detectection with PTZ camera
Sage Lab

PTZ App

Traffic Counting

2024

Code Generation and Documentation Search

Finding Events in Real Time

LIDAR for Solar Estimation and Sky Classification

Resource Management at the Edge

Sound Separation

Super Resolution Image Enhancement

2023 and earlier

AI@Edge for X-ray Beamline Experiments

Bandwidth Aware Learning

Characterizing Clouds

Lightning Detection w/ Software Defined Radio

Lightning Science

Monitoring biodiversity using acoustic data

Rideshare Sticker Detector

Snow Classifier

Snowflake Classification

Social Distancing

Solar Irradiance Estimation

Traffic State Estimation

Unleashing the Power of Collaboration

Vehicle Tracking

Water Level Detection

Water Segmentation

Wildfire Classifier

Use cases

dry landscape during drought
Use Case

Drought Mitigation and Prediction

thermal camera watching for wildfire
Use Case

Fire Science