Instrumentation Gym

AIS Instrumentation Gym

Welcome to AIS Instrumentation Gym

Follow the illustrated landscape from particles to planets, and from cells to data. Choose one of the seven NUS Science departments to discover its interactive demos.

Hover, focus, or tap a department to see its subtopics and current projects.

Introduction

Get to know the Gym

A two-minute guide to the Gym, its workflows, and how to get started.

AIS Gym / Introduction

Turn your phone sideways for the best view.

Project library

All interactive demos

Search the complete collection or narrow it by department.

Department
Method
Data Type
Instrument Type

TBB cluster explorer — grouping a day of Himawari-9 brightness temperatures

Physics Atmospheric Physics & Remote Sensing

TBB cluster explorer — grouping a day of Himawari-9 brightness temperatures

Can a computer rediscover a coastline it has never seen? A weather satellite photographed the heat of one patch of Japanese coast every ten minutes for a day — 142 pictures, two of the day's slots missing.

Clustering · PCA · UMAP · t-SNE · k-means
Open project
Pleiades membership — which stars move together?

Physics Astronomy & Astrophysics

Pleiades membership — which stars move together?

Can motion and distance separate the Pleiades from unrelated stars in the same patch of sky? A space telescope measured the position, motion, distance and light of thousands of stars around the Pleiades.

Clustering · Gaussian mixture model · HDBSCAN · k-means
Open project
Superconductor regression explorer — predicting critical temperature from composition features

Physics Condensed Matter & Materials Physics

Superconductor regression explorer — predicting critical temperature from composition features

Can a computer predict how cold a material has to be made before electricity flows through it with no loss at all? Twenty-one thousand materials that laboratories have already tested, each described by eighty-one numbers worked out from its chemical recipe, next to the temperature at which it was…

Regression · PCA · Linear regression · Polynomial regression · Random forest
Open project
Satellite Image Vector Quantization

Statistics and Data Science Machine Learning & Data Compression

Satellite Image Vector Quantization

See how clustering compresses twelve thousand measurements into a small codebook of representative patterns.

Clustering · k-means
Open project
Single Cell Gene Expression

Biological Sciences Genomics & Single-Cell Biology

Single Cell Gene Expression

Follow dimensionality reduction and clustering as twenty thousand gene measurements are organised into fourteen cell types.

Clustering · UMAP · k-means
Open project
Galaxy rotation-augmentation explorer — does turning the training images help?

Physics Astronomy & Astrophysics

Galaxy rotation-augmentation explorer — does turning the training images help?

Compare galaxy classifiers with and without rotation augmentation to see whether orientation-robust training improves predictions.

Classification · Convolutional neural network
Open project
Alzheimer's Gene Co-expression Explorer — the co-expression map, recomputed

Biological Sciences Genomics & Systems Biology

Alzheimer's Gene Co-expression Explorer — the co-expression map, recomputed

Explore gene co-expression patterns and clusters to identify biological structure associated with Alzheimer’s disease.

Clustering · PCA · t-SNE · k-means
Open project
JAE joint embedding — one cell, two measurements, 61 numbers

Biological Sciences Genomics & Single-Cell Biology

JAE joint embedding — one cell, two measurements, 61 numbers

Follow a joint autoencoder as one cell’s gene expression is compressed into a 61-number embedding.

Dimensionality reduction · Autoencoder
Open project
Air Quality — day-segment structure, sensor by sensor

Statistics and Data Science Multivariate Analysis & Visualisation

Air Quality — day-segment structure, sensor by sensor

Compare PCA and UMAP views of daily air-quality sensor profiles to reveal recurring patterns across sensors.

Dimensionality reduction · PCA · UMAP
Open project
Singapore Road Speed Clusters — UMAP

Statistics and Data Science Spatiotemporal Data Science

Singapore Road Speed Clusters — UMAP

Explore UMAP clusters of Singapore road-speed profiles to see which roads share similar traffic rhythms.

Clustering · UMAP · k-means
Open project
Battery SOH Forecast Explorer

Chemistry Electrochemistry & Energy Storage

Battery SOH Forecast Explorer

Can four ageing batteries reveal both how much capacity remains and how their voltage curves change? A laboratory repeatedly charged and discharged four matching batteries, recording one health value per cycle and the full voltage trace within each cycle.

Forecasting · LSTM
Open project
Battery Curve Shape Explorer

Chemistry Electrochemistry & Energy Storage

Battery Curve Shape Explorer

Can a computer recognise ageing from curve shape? Explore 3,245 cycles in one PCA map, reveal their ages, and inspect each linked charge-discharge curve.

Dimensionality reduction · Visualisation · PCA · UMAP
Open project
CEEMDAN battery forecasting — decompose, then forecast, without peeking

Chemistry Electrochemistry & Energy Storage

CEEMDAN battery forecasting — decompose, then forecast, without peeking

A rechargeable battery loses a little of the charge it can hold every time it is used. Can a computer see the end coming? One laboratory battery cell was charged and discharged 168 times, and the charge it could still hold was measured after every single cycle.

Forecasting · CEEMDAN · LSTM · Random forest
Open project