ES

Our Work

Portfolio

Engineering prototypes, software products, and technology solutions

Faena · Mining Analytics

18

The hub of browser-native mining apps: comminution, blasting, dispatch, scheduling, stockpiles, processing, geology and asset health, each with an offline precompute lane, a results ledger and a stated evidence boundary.

OreFlow, a Grinding and Separation Circuit Read as One Balance
Mining Analytics · Sep 2026

OreFlow, a Grinding and Separation Circuit Read as One Balance

In a processing plant a change made in one stage, a finer grind or a different classifier cut, moves every stage after it, and each area reports only its own part. OreFlow keeps the circuit as one balance: every stream carries the mass flow of every mineral in 63 size classes plus water, the grinding circuit is closed (an energy-specific population-balance ball mill with Plitt cyclones), separation runs through flotation banks, a gravity bleed, magnetic drums or desliming, and every unit closes within 1e-9. Twelve authored ore and plant scenarios with 96 variants; the workbench re-solves the circuit in the browser on every control change, and its optimizer and seeded uncertainty re-run there too, beside baked kinetic fits, Sobol sensitivity and five learned surrogates. 52 real ore samples and the hours of an iron-ore plant are workbench sources. Authored scenarios, not a calibrated plant.

mineral-processing grinding flotation mass-balance +6
Fragmenta, Blast Fragmentation Prediction Where the Learned Arms Explain Nothing on an Unseen Site
Mining & Optimization · Sep 2026

Fragmenta, Blast Fragmentation Prediction Where the Learned Arms Explain Nothing on an Unseen Site

Twelve arms predict the mean fragment size of a bench blast over sixteen cases and are scored under three ways of splitting the same table: six classical closed forms (Kuznetsov, Kuz-Ram, Swebrec, crush-zone, two published regressions) and six learned (a back-propagation network, support-vector regression, random forest, gradient boosting, a stack, a refit of the published regression). The result the product exists to report is negative: under leave-one-site-out not one of the six learned arms explains any variance, the best scoring -0.034 against a null model at -0.216, while the only two arms that stay positive on a site they have never seen are the two with fixed coefficients, the published regression at 0.802 and Kuznetsov at 0.311. The engine is the separately published blastfrag package. Lifecycle: building.

drill-and-blast fragmentation kuz-ram swebrec +6
TruckVitals, Onset Detection on Haul-Truck Telemetry (Regime First, Then Residual)
Predictive Maintenance · Aug 2026

TruckVitals, Onset Detection on Haul-Truck Telemetry (Regime First, Then Residual)

Every haul-truck channel moves with payload, grade and speed, so a change detector run on raw telemetry detects the truck going uphill, not a fault. TruckVitals segments the operating regime FIRST, from context channels that are disjoint from the monitored ones, and detects on the within-regime residual, so an alarm means a developing fault. The headline carries no detector in it: on six-operating-condition NASA C-MAPSS subsets the fault signature is 0.16 sigma of the pooled spread and 11.95 sigma of the within-regime spread, a ratio of 90.2, while the single-condition subsets return exactly 1.00 as a negative control nobody designed. It also ships a published NULL (regime conditioning does not improve onset localisation) and a WITHDRAWN claim (no false-alarm reduction is claimed anywhere, because the raw arm wins that metric on the synthetic lane). The headline changed after a review commissioned to refute it broke three of the four original numbers.

predictive-maintenance condition-monitoring change-point-detection regime-segmentation +4
PhaseFlow, Open-Pit Production Schedule with a Certified Bound
Mining & Optimization · Aug 2026

PhaseFlow, Open-Pit Production Schedule with a Certified Bound

An open-pit production schedule, solved with a certified bound and animated year by year over the block model. PitForge answers which blocks are worth mining; PhaseFlow answers when, subject to slope precedence in every period and per-period mining and processing capacity, maximising discounted NPV. That is the constrained pit limit problem (CPIT). The trust anchor is the published MineLib newman1.cpit solved AS PUBLISHED, on its own six periods, its own eight percent discount rate and its own two capacities: it reproduces the ultimate pit optimum exactly (26,086,899), its joint LP bound sits at 24,486,184, below the 24,486,549 LP bound published for the richer PCPSP problem, and its best schedule is 1.37 percent below that bound and 0.11 percent below an external integer optimum. The bound needs no LP solver (critical multiplier algorithm, exact in O(mn log n) as parametric maximum closures), which is why the whole problem re-solves live in the browser. Two bounds run on every case and the difference between them is reported, because that difference is the part of a gap that belongs to the bound rather than to the plan.

mining-optimization mine-planning scheduling cpit +5
StockTwin, Run-of-Mine Stockpile Twin (Built Truck by Truck)
category.mining-simulation · Aug 2026

StockTwin, Run-of-Mine Stockpile Twin (Built Truck by Truck)

A run-of-mine stockpile built the way one actually is: by trucks, one load at a time, following a dump plan, on ground they have to be able to climb. StockTwin routes every load over the trafficable surface, stands the material at its angle of repose, dozes the heaps into a floor, sorts each load down a real face with the Gray-Thornton kinetic-sieving model, and then reclaims in campaigns, so you can measure the three questions that matter: how much the pile homogenises the feed (variance reduction against the 1/N independent-layer bound, withheld where unreliable), where each reclaimed tonne came from (a per-cell lot ledger whose fractions sum to one on every cut), and how much size segregation biases the cut. Twenty-two scenarios across six axes, 22,656 placed loads, every one relaxing to ZERO cell pairs over the angle of repose, gated on the actual artifact. The physics is the separately published bedblend package; this repo declares none of its own. Synthetic and MineLib-public data only. Not a blending optimizer and it emits no plant setpoint.

mining-simulation stockpile blending grade-control +5
FrothSeg, Flotation-Froth Instance Segmentation Lab (Browser ML)
Computer Vision & AI · Jul 2026

FrothSeg, Flotation-Froth Instance Segmentation Lab (Browser ML)

A browser-native lab for instance segmentation of flotation froth: it delineates individual bubbles in a froth image, entirely client-side via ONNX, and compares a ladder of seven classical methods against a real published research model. That model, LamellaStar, is a four-head net shipped as a three-seed logit-mean ensemble (N1: mean AP 0.5186, AP50 0.8279, PQ 0.7359), with SAM2 and Cellpose-SAM as offline teachers and a distilled mask head for the browser. The central fact is the data situation, stated openly: no real froth images exist publicly (the search is recorded as a null, Roboflow dropped), so the froth cases are synthetic and BBBC038 (64 real dense-touching images, CC0) serves as an adjacent-domain transfer lane that the lab explicitly says does NOT clear the froth blocker. The study even refuted its own hypothesis: the froth ranking is generator-specific and the model drops on real transfer.

computer-vision instance-segmentation flotation froth +6
DispatchLab, Truck-to-Shovel Dispatch Bench for Open-Pit Mining
Mining & Optimization · Jul 2026

DispatchLab, Truck-to-Shovel Dispatch Bench for Open-Pit Mining

An in-browser truck-to-shovel dispatch bench: a deterministic discrete-event simulation of an open pit where you pick a case and a dispatch policy and watch tonnes, match factor and queues form. Nine policies (heuristic, OR-Hungarian, and learned) compete against a closed-form capacity oracle, over cases generated by minehaulsim, the author's own Apache-2.0 PyPI package. The distilled Monte-Carlo rollout is a published null result, winning 0 of 8 seeds under cycle-time uncertainty.

mining-optimization dispatch discrete-event-simulation open-pit +6
ChancaDEM, Crusher-Comminution Studio (Population-Balance Engine)
Mining & Optimization · Jul 2026

ChancaDEM, Crusher-Comminution Studio (Population-Balance Engine)

An in-browser crusher-comminution studio: you set the machine, closed-side setting, eccentric throw and speed and the feed gradation, and a pure-TypeScript closed-form engine (Whiten population balance + Evertsson capacity + Bond power) computes the product gradation, throughput and power in sub-millisecond. Despite the name it does NOT run DEM: the 3D chamber is a kinematic animation. The secondary-cone lane is calibrated against 10 published HP500 surveys (Rocha et al. 2024), with a leave-one-out cross-validation that includes negative controls.

mining-optimization comminution crushing population-balance +6
TailWatch, InSAR Ground-Deformation Studio for Tailings Dams & Slopes
Geotechnical & Risk · Jul 2026

TailWatch, InSAR Ground-Deformation Studio for Tailings Dams & Slopes

An in-browser InSAR ground-deformation studio for tailings dams and slopes. It renders a multi-temporal LOS-displacement cube (velocity, coherence, cumulative series) and runs classical velocity and inverse-velocity failure forecasting plus two small neural nets over it. Five cases are physics-simulated from a forward model; one is a real Sentinel-1 sample. On the held-out benchmark the classical velocity map (AUC 0.968) beats the learned anomaly autoencoder (AUC 0.898), and the app says so.

geotechnical insar ground-deformation tailings-dam +7
FragmentIQ, Post-Blast Muckpile Fragmentation Analysis Workbench
Mining Analytics · Jul 2026

FragmentIQ, Post-Blast Muckpile Fragmentation Analysis Workbench

A post-blast fragmentation workbench that delineates muckpile fragments by watershed in the browser and derives a mass-weighted particle-size distribution with a Rosin-Rammler fit and P10/P50/P80. It is scored against generator truth on synthetic muckpiles, with a real-photo lane of 5 CC BY images from an Iranian iron-ore mine where every number is explicitly RELATIVE, because no sieve ground truth exists, the scale is unknown, and n is small. The over-segmentation bias is shown directly in the artifact (116 recovered vs 70 true fragments), and the learned refinement gain is reported as indicative, not significant, on n=8.

mining-analytics fragmentation muckpile watershed +6
ProspectMap, Mineral Prospectivity Workbench with a Published Null Result
Mining Analytics · Jul 2026

ProspectMap, Mineral Prospectivity Workbench with a Published Null Result

A mineral-prospectivity workbench that computes a Weights-of-Evidence posterior P(deposit|evidence) live in the browser over stacked geophysical, geochemical and structural layers, and whose real reason to exist is to expose how prospectivity maps overstate their skill. On real US Midcontinent MVT Zn-Pb belt data (Lawley et al. 2022, USGS public domain) it publishes a recorded null: the proposed PU-Conformal method loses to 1989-vintage Weights of Evidence, and a trivial distance-to-deposit baseline already ties the best learned model, so most apparent skill is spatial proximity, not learned geology. It makes first-class the two ways prospectivity maps lie: conditional-independence violation and random-CV inflation.

mining-analytics mineral-prospectivity weights-of-evidence null-result +6
PitForge, Exact Ultimate Pit Limit & Nested Whittle Pit-Shell Workbench
Mining & Optimization · Jul 2026

PitForge, Exact Ultimate Pit Limit & Nested Whittle Pit-Shell Workbench

An open-pit mine-design workbench that solves the ultimate pit limit exactly, as a maximum-weight closure of the block-precedence graph reduced to a minimum cut on a Dinic max-flow engine, running live in the browser, and derives the nested Whittle pit shells by revenue factor. It reproduces the published optima of three real MineLib instances to at most 2e-9 relative error, and self-checks every solve with the max-flow duality identity. This is the only one of the Faena four whose headline is validated against third-party published optima rather than its own generator.

mining-optimization open-pit ultimate-pit-limit max-flow +6
CoreLog Vision, Drill-Core Logging & Lithology Classification Workbench
Mining Analytics · Jul 2026

CoreLog Vision, Drill-Core Logging & Lithology Classification Workbench

A browser-native drill-core logging workbench: it classifies sliding windows along core-tray channels into six lithologies with a CNN, merges adjacent same-class patches into a depth-stitched strip log with confidence shading, and flags out-of-distribution core with a Mahalanobis detector instead of forcing a class. It runs on procedurally synthetic trays, with a real-photo lane (DCID) used strictly as the out-of-distribution and real-head evaluation set. The CNN accuracy is reported as synthetic-vs-synthetic, a label-permutation control collapses to chance, and the shipped detector is named for what it is.

mining-analytics drill-core lithology cnn +6
MineHaul, Mine-Haulage Discrete-Event Simulation
Simulation & Operations Research · Jul 2026

MineHaul, Mine-Haulage Discrete-Event Simulation

An open-source Python package (minehaulsim) for deterministic discrete-event simulation of open-pit and underground mine haulage on constrained road networks, with seeded parametric mine generators. Byte-deterministic, numpy-only, well-tested, the companion generator that feeds structure-real scenarios to DispatchLab.

simulation des mining haulage +3
CutoffGrade Studio, Lane's Optimal Cut-off Grade
Mining & Optimization · Jun 2026

CutoffGrade Studio, Lane's Optimal Cut-off Grade

An open, explainable studio for Lane's optimal cut-off grade: feed a grade-tonnage curve plus prices, costs and three stage capacities, and it computes the NPV-maximising declining cut-off trajectory, NPV, mine life and cashflow, live in the browser, with the exact algorithm as the authority and a learned surrogate only for speed.

mining optimization economics cutoff-grade +3
Faena, Mining-Analytics Hub
Mining Analytics · Jun 2026

Faena, Mining-Analytics Hub

A single open launcher that catalogues a growing set of independent, in-browser mining-analytics tools across the value chain: exploration, drill & blast, load/haul, comminution, processing, tailings, asset health and economics. Each tool is its own documented product on a named real dataset or a validated synthetic. Six live today, eight in active development, and 28 more mapped on a visible roadmap.

mining analytics hub launcher +2
RotorVitals, Rotating-Machinery Condition Monitoring & Prognostics
Predictive Maintenance · Jun 2026

RotorVitals, Rotating-Machinery Condition Monitoring & Prognostics

An in-browser condition-monitoring and prognostics workbench for rotating machinery (bearings-first), running on real measured vibration. A source selector switches between a synthetic signal generator, real diagnosis segments (CWRU / Ottawa order-tracked / MaFaulDa), and real run-to-failure trajectories (FEMTO / XJTU / IMS). A classical DSP chain, a learned WDCNN + deep-autoencoder, and a four-model RUL ladder run live via onnxruntime-web. Synthetic data is labelled as such, cross-domain transfer is flagged, and frequency relations are exact.

predictive-maintenance condition-monitoring prognostics rul +7
ChargeCascade, Tumbling-Mill Charge Motion & Power Studio
Mining & Optimization · Jun 2026

ChargeCascade, Tumbling-Mill Charge Motion & Power Studio

A live in-browser 3D studio for tumbling-mill (SAG / ball / rod) charge motion and power: critical speed, the Davis cascading → cataracting → centrifuging transition, toe/shoulder angles, and Hogg-Fuerstenau / Morrell-form / Bond power, published closed-form physics recomputed on every slider, with a trained ONNX power surrogate and an out-of-envelope anomaly guard running client-side.

mining comminution simulation 3d-visualization +3

Mining Operations & Processing

7

Mining products outside the hub: ventilation, drillholes, stockpile visualization, geotechnical risk, quantum optimization measured on mining problems, and phenomenological and learned models of mineral processing.

Fragua, Ensembles of Phenomenological Models for Mining and Industrial Processes Proprietary
Scientific Machine Learning · Sep 2026

Fragua, Ensembles of Phenomenological Models for Mining and Industrial Processes

Instead of selecting one kinetic equation per unit process, Fragua fits many realizations, model families times parameter multistarts times bootstrap resamples, from a curated bank of phenomenological families, and aggregates them into calibrated ensemble models with structural inclusion probabilities: which model describes this process, with what confidence, and what that structural uncertainty means for prediction beyond the training envelope. The coined method, BAPE (Bootstrap-Aggregated Phenomenological Ensembles), is rung 6 of a twelve-rung measured ladder that also carries controls, a Bayesian model average, sparse discovery, a Kennedy-O'Hagan hybrid and deep ensembles, baked canonically over 30 variants; the browser runs a live lane through the published phenoforge wheel. A private research product with the concept's novelty gap verified against literature and vendors. Lifecycle: building.

phenomenological-models ensembles bootstrap model-selection +4
Aerovia, an Underground Ventilation Design and Analysis Workbench in the Browser
category.mining-simulation · Sep 2026

Aerovia, an Underground Ventilation Design and Analysis Workbench in the Browser

Draw a ventilation network in 3D, place boundaries and fans, and solve it: flow, pressure, paths, recirculation, fan duty, sensitivity, energy and baseline comparisons, then release a conservative passive tracer with pulse or continuous sources, scheduled changes, timeline playback and a complete mass ledger. Trained MLP and graph models ship as 24 ONNX exports with their held-out approximation limits stated. Twelve authored network cases, 33,792 learned dataset states and a 216-cell benchmark are reproducible from the public scripts; imports, numerical workers and inference all run in the browser. The networks are planning inputs, not measured operational mines, and no industrial saving or field calibration is claimed. A first version was rejected and rebuilt in full from the CAOS template.

mining ventilation network-solver tracer-transport +4
Sondara, Drillhole Estimation and Simulation Compared on Real Held-Out Holes
Geotechnical & Risk · Sep 2026

Sondara, Drillhole Estimation and Simulation Compared on Real Held-Out Holes

Three public drillhole families run through one reproducible pipeline, and twelve estimation and simulation methods, from nearest neighbour and kriging to SNESIM, Direct Sampling, DeepKriging and KCN, predict the same held-out holes under one scoring. On Rocklea iron ordinary kriging holds its place; the learned methods learn the structure without beating it.

drillholes geostatistics mining kriging +4
Porvenir, Learned World Models for Mineral Processing, Measured Against Holding the Last Value Proprietary
Scientific Machine Learning · Sep 2026

Porvenir, Learned World Models for Mineral Processing, Measured Against Holding the Last Value

A forecaster maps history to a future; a world model takes a proposed action sequence as an argument and returns a distribution over futures, which is the single difference that makes what-if questions, planning and regime detection expressible. Porvenir asks three questions of mineral-processing records and answers each with a measurement: whether a learned latent can track process state the plant does not instrument, whether aleatoric and epistemic uncertainty can be separated and made separately actionable, and how far imagination drifts from reality when a plan is executed. Eleven rungs plus a zero-shot yardstick, from persistence and seasonal naive through VARX, GRU-D, DeepAR, MQ-RNN, a probabilistic ensemble, a patch transformer and recurrent state-space models, every one scored per horizon against holding the last value. On the real iron-ore flotation record, seven of nine learned rungs do not beat persistence, and the app says so on the case. Lifecycle: live, re-baked with its numbers withdrawn once.

world-models forecasting latent-dynamics uncertainty +5
QMine, Quantum Computing Measured on Real Mining Problems Proprietary
Quantum Computing · Aug 2026

QMine, Quantum Computing Measured on Real Mining Problems

A private research app that measures what today's quantum providers can actually do on real, data-driven mining problems. Every case runs the same ladder end to end: proven-optimal classical solvers, the field heuristics a plant uses, the strongest quantum-inspired algorithms, the gate-model method family on simulators, and the real provider lanes when their credentials exist, with a money-denominated cost ledger and a computed verdict on every instance. Sixteen cases, 121 variants and 3,255 declared solver cells across six categories plus a flotation soft-sensor prediction problem, baked completely with every skip explicit. The honesty spine, on screen per case: as of the research date there is no third-party-confirmed quantum advantage for any industrial optimization problem, and today the classical baseline usually wins. Lifecycle: building.

quantum-computing optimization qaoa annealing +5
Mine Pile Visualizer, Stockpile & Circuit Topology Explorer
3D & Visualization · Mar 2026

Mine Pile Visualizer, Stockpile & Circuit Topology Explorer

A local-first web app for exploring mining circuit topology, live material state, stockpile 3D structure, and historical profiler data. Built on Next.js 16 with React Three Fiber.

mining visualization nextjs react +4
Geotechnical & Risk · Mar 2026

UnderMine Risk, Geotechnical Risk Operator Dashboard

An interactive operator dashboard that turns weekly geotechnical risk model output into an actionable mine map. Per-extraction-point risk on a deck.gl view, aggregation layers, per-lever actionables, and printable monthly reports. A portfolio demo running on fully synthetic data.

geotechnical mining-safety deckgl nextjs +3

Earth Science & Geospatial 3D

9

Seismic forecasting, geophysical inversion, fractured-reservoir well tests, geospatial tooling, real places rebuilt in 3D, streaming reconstruction, and information-theoretic sampling of geological images.

Inverse Earth Studio, a Geophysical Inversion Workbench on Original Synthetic Earths
Geophysics & Seismic Hazard · Sep 2026

Inverse Earth Studio, a Geophysical Inversion Workbench on Original Synthetic Earths

A bilingual geophysics research workbench covering 3D potential fields, layered magnetotellurics, acoustic full-waveform inversion, learned inverses and joint cross-gradient inversion, on twenty distinct geological constructors with six computed conditions each: 120 artifacts and 348 method results, replayed in the browser with an inversion scrubber, a 3D scene with cut, threshold and rotation, and live MT parity. Fifteen numerical tests include an executed Deepwave CUDA adjoint derivative, independent Choclo and SimPEG prism checks, complex MT parity and checkpoint reload; the learned split holds 800, 160 and 160 independent realizations. Original synthetic examples, not field validation; no posterior uncertainty or algorithmic novelty is claimed. Version 2 replaced a first version its owner rejected.

geophysics inversion gravity magnetotellurics +6
Rajo, Open Pits Seen from Orbit
3D & Visualization · Sep 2026

Rajo, Open Pits Seen from Orbit

An open observatory of the world's great open-pit mines and lithium evaporation ponds over four decades: a 3D globe with real relief, a yearly time-lapse per site from Landsat 1985 to Sentinel-2 today, spectral and mineral indices computed live in the browser from the raw cloud-optimized GeoTIFFs, a learned mine-footprint segmentation running in the browser, change points on the mined-area signal, and the elevation difference between the 2000 and the 2011 to 2015 global surfaces. Thirty sites, Chilean copper at the core; no backend, no account, nothing uploaded.

remote-sensing landsat sentinel-2 mining +9
GeoLab, Browser-Native Multi-Engine Geospatial Tool Platform
Data Engineering · Jul 2026

GeoLab, Browser-Native Multi-Engine Geospatial Tool Platform

A browser-native, no-install geospatial platform: load a DEM, GeoTIFF, vector or point cloud (a bundled sample or your own file), run real geoprocessing tools, chain them into reproducible pipelines, and explore the result on a map or 3D canvas, entirely in your browser. No server, no Python, no GDAL install, and your data never leaves your machine. Today it runs 747 real WhiteboxTools / GeoLibre tools client-side via the geolibre WASM engine; the architecture (ADR-0059) is multi-engine by design, aggregating GDAL, GEOS, Turf, H3, mapshaper, ITK-Wasm, OpenCV.js and ONNX behind one uniform tool interface as adapters land incrementally. Every tool shows its source engine, authors and license, and the tool count is only what genuinely runs, never padded.

data-engineering geospatial gis webassembly +5
Maqueta, Real Places Rebuilt in 3D from Fused Open Geodata
3D & Visualization · Jul 2026

Maqueta, Real Places Rebuilt in 3D from Fused Open Geodata

Take a real area and rebuild it in 3D from public, open geodata, with the height map as the way in. Maqueta fuses Overture buildings and roads, Copernicus GLO-30 terrain, ESA WorldCover land cover, OpenStreetMap water, green and rail, GHS-POP population, Google Open Buildings 2.5D heights, 3DBAG LoD2 as an official reference layer and SoilGrids soil carbon; every layer keeps its source and licence visible, and every building height carries its provenance: measured, inferred from floor count, read from a height raster, or a default. 118 places baked and served, from cities on every continent to Santiago split into its 37 comunas as separate cases. Colour and filter buildings by any fused attribute, click one to read everything fused into it, drape a cloudless Sentinel-2 image, and draw a polygon over any sub-area to summarise the buildings inside: count, footprint area, built coverage, density, height distribution, floors, and the mix of function, land cover and height provenance. The fusion engine is the separately published geoscena package.

geospatial 3d-visualization data-fusion open-data +6
Pulso, Well-Test Diagnostic-Curve Shape Clustering and Attribution
Scientific Machine Learning · Jul 2026

Pulso, Well-Test Diagnostic-Curve Shape Clustering and Attribution

An unsupervised catalogue of flow-behaviour classes ("GeoTypes") for fractured reservoirs: it clusters the SHAPE of pressure-transient Bourdet-derivative curves with DTW k-medoids, attributes each class to the fracture-network descriptors that control it (Random Forest + SHAP), and ships a browser workbench that classifies a user curve against the baked medoids with conformal prediction. It reproduces and extends Kamel Targhi et al. 2026 (Comp. Geosciences 30, 57) on a real 4TU corpus (~4768 curves), then transfers the same shape diagnostic to two real aquifer pumping-test sites, with an explicit caveat that the shape transfers but the physics does not.

scientific-ml well-test pressure-transient bourdet +8
Lidar3D, Streaming 3D Reconstruction Lab
3D & Visualization · Jul 2026

Lidar3D, Streaming 3D Reconstruction Lab

An early research lab for feed-forward streaming 3D reconstruction: an ordered RGB or LiDAR stream becomes a camera trajectory, dense metric depth and a fused colored point cloud. The heavy engine runs offline on a GPU; the site replays the baked cloud across four point-cloud renderers under one unified transform, with strict renderer parity as the point: every renderer draws the same data, and approximations are disclosed.

3d-visualization research computer-vision point-cloud +2
CAOS Seismic, Probabilistic Earthquake Forecasting
Geophysics & Seismic Hazard · Jun 2026

CAOS Seismic, Probabilistic Earthquake Forecasting

Earthquakes cannot be predicted, but their probability can be forecast. A daily, calibrated, conditional probability of seismic events over 1/2/7-day horizons, always shown next to its long-term baseline and scored prospectively against reality with the field-standard CSEP framework. A forecaster, never an alarm.

seismology forecasting probability etas +4
Information-Theoretic Sampling for Geological Image Recovery
Research & Thesis · Dec 2019

Information-Theoretic Sampling for Geological Image Recovery

Doctoral thesis addressing optimal sensor placement, where to place N measurements to minimize posterior uncertainty in binary random fields. Introduced AdSEMES algorithm with submodularity guarantees. Published in Mathematical Geosciences and Natural Resources Research.

phd-thesis information-theory entropy geostatistics +2
Optimal Spatial Sampling via Information Theory
Research & Thesis · Feb 2016

Optimal Spatial Sampling via Information Theory

Implements the AdSEMES algorithm for optimal spatial sampling in binary random fields, specifically channelized geological reservoirs. Companion code to published papers in Mathematical Geosciences and Natural Resources Research.

information-theory entropy sampling geostatistics +2

Physics, Mathematics & Scientific Computing

6

Research labs on a method family or a theorem, end to end: optimal control of magnetization, the three-gap theorem, symbolic regression, simulation, physics-informed networks and quantum problems on simulators, with published negative results.

GapTheo, a Visual Atlas of the Three-Gap Theorem
Research & Thesis · Sep 2026

GapTheo, a Visual Atlas of the Three-Gap Theorem

Place the first N multiples of an irrational number on a circle and the gaps between neighbours take at most three distinct lengths. GapTheo keeps that sorted finite orbit as the primary live calculation and synchronises it with eleven linked readings of the same object: Farey-cell geometry, continued fractions, return gaps, cyclic words, a lattice schematic, a two-interval exchange lens and a finite 0D Rips topology view. Twelve canonical cases, an exact finite direct oracle, generated certificates, a staged reference pipeline, a complete benchmark, the manuscript source and a deep wiki ship together. Seeded random placement, a farthest-point allocator and a two-frequency extension are explicit contrast regimes, never silently treated as inputs to the classical theorem. No new theorem or formal proof is claimed.

mathematics dynamical-systems three-gap-theorem continued-fractions +4
Espira, Optimal Control of Magnetization Switching in Two-Dimensional Magnets
Simulation & Operations Research · Sep 2026

Espira, Optimal Control of Magnetization Switching in Two-Dimensional Magnets

Given a magnetic bit in a two-dimensional van der Waals magnet and a target switching time, compute the field or current pulse that flips it for the least dissipated energy. Espira reproduces the kickoff paper (Badarneh, Cai and Santos, Adv. Mater. 2026, doi 10.1002/adma.202523059) as a verification floor and extends it: analytic optimal paths for uniaxial and spin-orbit-torque switching, a numerical optimal-control problem for the biaxial case, conventional baselines, GRAPE and CRAB, a reliability front, and the beyond-macrospin question answered on a free chain, where domain walls become the optimal reversal above a crossover length. Two independent codes agree with the engine: Spirit on the barrier to 1.4e-06 over six lattices, VAMPIRE on the equation of motion to 4.7e-06, both crossing the equator at 78.7319 ps. 25 of 26 cases baked; two run live in the browser. The engine is the separately published spinoct package.

optimal-control spintronics magnetization-switching van-der-waals +5
SymLab, Symbolic Regression Lab (Accuracy and Recovery, Reported Separately)
Scientific Machine Learning · Jul 2026

SymLab, Symbolic Regression Lab (Accuracy and Recovery, Reported Separately)

A public research lab on symbolic regression: recovering an explicit closed-form expression from data, rather than fitting a predictor nobody can read. Its reason to exist is one distinction: accuracy and structural recovery are different claims, and a method can clear R2 above 0.999 while returning a structure with nothing in common with the law that generated the data. SymLab reports the two separately, on every case, and never averages them into one number. It runs multiple search families (genetic programming with linear scaling and deduplication, sparse regression) at published budgets, live in the browser, on benchmark laws where the true expression is known.

scientific-ml symbolic-regression genetic-programming sparse-regression +2
PINN-Lab, A Runnable Catalogue of Physics-Informed Neural Networks
Scientific Machine Learning · Jun 2026

PINN-Lab, A Runnable Catalogue of Physics-Informed Neural Networks

A live catalogue of 19 Physics-Informed Neural Network cases: each trained offline (DeepXDE → ONNX), validated against an analytic, benchmark, or real-data anchor, and re-inferred live in the browser: move a physical parameter and watch the trained network re-solve the PDE.

scientific-ml pinn deepxde onnx +3
QLab, Quantum Laboratory
Quantum Computing · Jun 2026

QLab, Quantum Laboratory

A public, didactic quantum-computing lab that runs real frameworks (Qiskit, PennyLane, Cirq, Stim) on 20 worked cases and puts every quantum method next to its classical baseline, and shows that at lab scale, classical still wins.

quantum-computing qiskit pennylane cirq +3
CAOS SimLab, A Didactic Lab for Simulation (DES & ABM)
Simulation & Operations Research · Jun 2026

CAOS SimLab, A Didactic Lab for Simulation (DES & ABM)

Land straight in a running simulation, move the sliders, and watch the dynamics change. An open lab for Discrete-Event Simulation and Agent-Based Modeling: ten worked scenarios across six visualization families, each a pure function of (params, seed) so replay is exact, validated against theory where it exists.

simulation des abm operations-research +4

Life Sciences & Neuroscience

5

Cardiac inverse problems, cell migration and Potts models, and the fly connectome used as a frozen network and driven as a whole nervous system, every claim measured against null controls.

Destello, the Whole Fly Nervous System Working Through Its Two Real Eyes
Scientific Machine Learning · Sep 2026

Destello, the Whole Fly Nervous System Working Through Its Two Real Eyes

The male fruit fly's whole nervous system, working, seen through its two eyes: the MaleCNS v1.0 connectome (166,700 neurons, brain and nerve cord, CC BY 4.0) driven through both real compound eyes (879 left and 892 right measured columns) and run with the published neuron models, a graded optic lobe and a leaky integrate-and-fire model elsewhere, so a visitor can watch real neurons fire as real soma positions and real arbors lighting up, from the photoreceptors to the descending neurons and the nerve cord, in slow motion with the biological clock on screen, and intervene. The science: which visual computations and behaviours the measured wiring produces on its own, against three null wirings, and whether depth and fast segmentation can be read from the real neurons whose job it is. State: the engine, flycns, is complete in Python (0.06.000) and measured on the real connectome; the web product is not built and nothing is deployed. Lifecycle: planned.

connectome drosophila neuroscience simulation +5
Conectoma, a Fly Connectome Used as a Frozen Vision Network, Measured Against Its Own Nulls
Scientific Machine Learning · Sep 2026

Conectoma, a Fly Connectome Used as a Frozen Vision Network, Measured Against Its Own Nulls

The Janelia MaleCNS v1.0 wiring diagram of the male fruit fly (166,691 neurons, 11,691 cell types, CC-BY) frozen and used as the architecture of a computer-vision network, to measure what a network built from that biology can compute: depth from a moving camera and figure-ground segmentation. Synapse counts and signs never change; three regimes differ only in what the optimizer may touch, and every result is an effect size against null controls built from the same wiring: degree-preserving rewiring, a size-matched random sparse graph, a sign shuffle. With the connectome as a reservoir and a trained readout, paired per clip: +0.197 against the random sparse graph, +0.136 against the sign shuffle, and -0.017 against the degree-preserving rewiring. Live at 0.10.001 with the pathway animated; its plan is not yet validated and the lifecycle reads planned.

connectome drosophila computer-vision depth-estimation +5
CardioPINN, Physics-Informed Cardiac Inverse-Problem Lab
Scientific Machine Learning · Jul 2026

CardioPINN, Physics-Informed Cardiac Inverse-Problem Lab

A two-case lab in physics-informed cardiac reconstruction, on 100% real measured data, scoped as a complement to classical methods, not a replacement. Case 1 (ECGi) recovers heart-surface potentials from body-surface potentials with regularized least squares plus a graph prior and a deep ensemble; case 2 (4D-flow) recovers the aortic relative-pressure FIELD from MRI velocity with a genuine divergence-free PINN. What the physics adds is what the classical method cannot give: calibrated per-node uncertainty (2-sigma coverage ~0.90) and a resolved pressure field from a well-posed solve, a different output class than a one-number Bernoulli estimate. Point accuracy sits at parity with a well-tuned Tikhonov baseline, stated in-app, because replacing classical accuracy was never the goal. Two confirmed, CI-tested autograd advances (spatial and temporal) and three published null results.

scientific-ml physics-informed pinn inverse-problems +6
Cellular Potts Model Simulator
Biophysics & Simulation · May 2013

Cellular Potts Model Simulator

A CPM simulator for modeling collective migration of Dorsal Forerunner Cells during zebrafish development. Cells as deformable bodies on a lattice with Gaussian filopodia, durotaxis, and contact interactions at 10-50 FPS.

biophysics cellular-potts-model simulation computational-biology +2
3D Embryo Cell Migration Simulator
Biophysics & Simulation · Aug 2011

3D Embryo Cell Migration Simulator

A 3D simulation of Deep Forming Cell collective migration on the surface of a spherical zebrafish embryo during epiboly. Uses AER coordinate system with WebSocket streaming and Three.js visualization.

biophysics simulation three-js fastapi +2

Optics, Vision & Imaging

9

Super-resolution microscopy, holographic tweezers, dual photography, hyperspectral geometallurgy, damage vision in engineering units, RGB-D measurement and the mathematical representations of an image.

Fisura, Materials-Damage Vision Lab (Crack Detection to Engineering Units)
Computer Vision & AI · Jul 2026

Fisura, Materials-Damage Vision Lab (Crack Detection to Engineering Units)

A public research lab on seeing damage in built materials. One image of a concrete wall, pavement, masonry facade or industrial surface goes in; Fisura detects the damage (cracks, spalling, surface defects), quantifies it in engineering units (width, length, orientation, density, growth between inspections), and shows how every method family gets there on the same open cases with the same metrics. The ladder runs classical pipelines, learned SOTA segmentation, promptable foundation models and unsupervised anomaly detection, then measurement (pixel-to-mm calibrated width/length, change detection, 2D digital image correlation). Masks are never the end result: they are the input to engineering numbers. Offline heavy lane plus a browser live lane where the photo never leaves the device. Scope: it is a method-comparison lab, not a certified inspection tool, and the classical engine is real while the higher tiers land one vertical slice at a time.

computer-vision crack-detection materials segmentation +5
ImageLab, One Image Across the Full Spectrum of Mathematical Representations
Image Representation · Jul 2026

ImageLab, One Image Across the Full Spectrum of Mathematical Representations

One image written across eleven mathematical representations: orthonormal transforms (Fourier, DCT, wavelet, KLT), overcomplete sparse dictionaries, geometric primitives, an implicit neural field (SIREN), symbolic CPPN formula art, Fourier-descriptor epicycles, and learned generative latents (VAE, diffusion), each with editable parameters. The measured thesis is that editability is U-shaped: it peaks at the designed-structure pole (local, exact edits) and the learned-manifold pole (semantic, entangled edits) and collapses to noise between. Everything is computed live in the browser (TypeScript, WebGL2) or baked offline by an open, seed-deterministic Python pipeline.

image-representation signal-processing fourier wavelet +6
LDA-HSI, Wordification Design-Space Platform
Spectral Analysis · May 2026

LDA-HSI, Wordification Design-Space Platform

A research platform that asks which representation a hyperspectral topic model should see: 19 wordification recipes × 4 topic-model backbones × Q∈{8,16,32}, scored on a 12-axis battery across six public scenes, with a live web app and API.

hyperspectral topic-modelling representation-learning lda +3
Hyperspectral Mineral Analysis by Topic Modelling (2022)
Spectral Analysis · Aug 2022

Hyperspectral Mineral Analysis by Topic Modelling (2022)

The original 2022 conference work that reframed hyperspectral mineral characterisation as a topic-modelling problem: spectra as documents, LDA topics as a routing layer for per-topic regression. Cut copper-recovery error an order of magnitude on private drill-core data.

hyperspectral topic-modelling lda geometallurgy +2
Dual Photography Lab
Computational Optics · May 2020

Dual Photography Lab

An interactive application implementing the dual photography technique: reconstructing how a scene looks from a projector's point of view using light transport matrices, SVD decomposition, and Helmholtz reciprocity.

computational-photography linear-algebra svd compressed-sensing +2
GrainSight, Particle Size Distribution from RGB-D Data
3D & Visualization · May 2018

GrainSight, Particle Size Distribution from RGB-D Data

A 3D particle size and granulometry analyzer using marker-based watershed segmentation on RGB-D data. Extracts 18 geometric properties per grain and fits Rosin-Rammler PSD distributions with ISO 565 sieve simulation.

granulometry particle-size watershed depth-profiling +3
SurfaceScope, RGB-D Depth Profiling for Surface Analysis
3D & Visualization · Jul 2017

SurfaceScope, RGB-D Depth Profiling for Surface Analysis

An RGB-D depth profiling system for analyzing mineral sample surfaces. Creates 3D reconstructions via pinhole camera model with bilateral filtering, normal estimation, curvature computation, and ISO 4287 roughness metrics.

depth-profiling 3d-reconstruction surface-analysis granulometry +3
Super-Resolution Microscopy via Optical Fluctuation Imaging
Computational Optics · Jun 2012

Super-Resolution Microscopy via Optical Fluctuation Imaging

Implements SOFI, a computational super-resolution technique extracting sub-diffraction spatial information from temporal fluorescence fluctuations of quantum dots. First successful SOFI implementation in Chile. Resolves ~120 nm features below the 232 nm diffraction limit.

super-resolution microscopy sofi cumulants +3
Dynamic Holographic Optical Tweezers
Computational Optics · Mar 2010

Dynamic Holographic Optical Tweezers

Real-time computation and visualization of holographic phase masks for optical tweezers systems. Implements the weighted Gerchberg-Saxton algorithm for shaping laser beams into multiple focused traps to manipulate microscopic particles.

optics holography computational-physics fastapi +1

AI, Agents & Software Systems

8

Agent-based systems that must stay accountable: adversarial review, a single-seeker career workbench, a management console, industrial agent hubs, learned metacontrol, and the measurement of language models turning statements into formal models.

Neuraxis and reflexmesh, an Event-Driven Software Controller with Learned Metacontrol Proprietary
Agentic AI & Automation · Sep 2026

Neuraxis and reflexmesh, an Event-Driven Software Controller with Learned Metacontrol

A fast, typed controller for software, files, tools and agents, with the ability to recruit slower language-model reasoning only where the outcome justifies it. The reusable core, reflexmesh, is a public Rust/PyO3 and Python package on PyPI: typed event-driven admission, registered fast actions with verified effects, learned control. Neuraxis is the private app on top: three native file tasks with real findings (verify a software release, reconcile CSV records, audit sources), matched controller executions that expose the controller's evidence, and the scientific pages with shared tabs and a bibliography. The research claim is outcome-grounded allocation of deliberation in a recurrent, resource-versioned asynchronous controller. The corrected evaluation, 19,488 episodes, does not establish learned superiority: the fitted allocation gates select never-defer, and four fixed gates match the learned one. A harder semantic-transfer lane of 1,536 episodes, built so that a fixed controller cannot solve everything, does not establish it either: the always-planner method verified 38 of 128 episodes and the learned controller 18. No AGI or biological claim is made.

software-control agent-systems learned-metacontrol rust +4
Enunciado, How Faithfully Language Models Turn a Problem Statement into a Solvable Model
Scientific Machine Learning · Sep 2026

Enunciado, How Faithfully Language Models Turn a Problem Statement into a Solvable Model

The field reports whether a generated model ran and calls that correct. Enunciado measures faithfulness instead: a corpus of 20 authored optimization statements across five complexity tiers, every trap covered and four controls without one, is handed to language models, and the formal model each returns is judged by an oracle that is not a language model: executable, structural and property layers over a solver, with a duality certificate. Sixteen models from four providers, 320 calls, every row complete. The best local model, phi4, is faithful on 5 of 20. The output cap decides the reasoning models: DeepSeek-V4-Pro is faithful on 2 of 20 at 8192 tokens and 11 of 20 at 32768. The first version of this measurement read +0.000 and was wrong.

formalization optimization autoformalization llm-evaluation +5
CareerTwin, a Single-Seeker Career Evidence and Job-Search Workbench
Agentic AI & Automation · Sep 2026

CareerTwin, a Single-Seeker Career Evidence and Job-Search Workbench

A self-hosted workbench where one seeker owns one evidence-centred profile and any number of opportunities, applications, tasks and generated career artifacts. Its match score is a versioned alignment measure with coverage and uncertainty, never a hiring probability: it separates required, preferred and eligibility requirements, states how much of each is covered by confirmed evidence, and explains every component reproducibly. The agent side is bounded and evidence-cited, every proposed canonical write needs explicit approval, and the whole product works without any AI provider. It does not rank candidates for employers, infer protected traits, scrape unrestricted job sites, auto-apply or send outreach. Live at v0.14.5 on a single VPS; no external job-search outcome is claimed.

career evidence-graph job-search matching +7
CentralIA, Initiative-Management Console over a Version-Control Provider
Data Engineering · Jul 2026

CentralIA, Initiative-Management Console over a Version-Control Provider

A connection-first web console that connects to a version-control provider (GitHub, GitLab, or an offline emulator), reads a management repo through its structure map, and renders a whole portfolio of repo-mapped initiatives: a filterable grid, per-initiative 5-axis status (development, design, implementation, deployment, value), implementation backlog, live open issues, plans, findings and history, plus a hub drill-down with a growth graph showing how a suite of components knits together through consumes edges. It is a real tool, not a mockup: the connectors are genuinely implemented and the offline emulator is a self-contained backend with a bundled fake management repo. The access token lives in the server environment, never the browser or a database, and a server-side anonymize toggle redacts client info for screen-sharing.

data-engineering developer-tools project-management github +4
ADRA, Adversarial Dev Review Agent
Agentic AI & Automation · Jul 2026

ADRA, Adversarial Dev Review Agent

A deterministic-first, adversarial-validation engine for the software lifecycle. Deterministic tools (git, the exact CI command, bundle validation, a language/leak scan, SQL probes) run FIRST and become both the grounding the model may not contradict and the evidence in an immutable provenance log; a blocking adversarial critic tries to refute each artifact instead of blessing it, an LLM-as-judge scores with swap-and-average, and the loop escalates to a human where nothing deterministic backs the verdict. Six skills (code review, PR eval, experiment, improve, document, decide) run the same loop. Runs offline with no API key; published on PyPI as adra.

agentic-ai code-review adversarial-validation llm-as-judge +5
Mantia, Anonymized Hub of Agentic Tools for SAP Plant Maintenance Proprietary
Agentic AI & Automation · Jun 2026

Mantia, Anonymized Hub of Agentic Tools for SAP Plant Maintenance

An anonymized hub cataloguing a family of agentic tools oriented to SAP Plant Maintenance (PM). Each tool automates a maintenance workflow (work-order generation, spares planning, notification triage, scheduling and more) on one shared agent core with a mock-or-real connector to the enterprise system. Delivered under confidentiality: general description only, no client data.

agentic-ai maintenance odata eam +3
Veta, Agentic Mineral-Tracking Decision Support Proprietary
Mining & Optimization · Jun 2026

Veta, Agentic Mineral-Tracking Decision Support

A private decision-support system that turns a natural-language question (typed or spoken) about a mineral-processing operation into a traceable, evidence-backed answer, a multi-stage agent pipeline routes the query to tiered solvers and grounds every recommendation in plant data with RAG.

mining decision-support agents rag +3
Circuita, Stochastic Mineral-Tracking Engine & Console Proprietary
Mining & Optimization · Jun 2026

Circuita, Stochastic Mineral-Tracking Engine & Console

A private analysis console built on a stochastic cellular-automaton engine that tracks material through a processing circuit. It produces fan, balance, delta and inventory views per pile and exports a self-contained, bilingual report with inline-SVG charts.

mining cellular-automaton simulation material-tracking +2

Data, Signals & Forecasting

4

A forecasting atlas with a foundation-model tier, quantitative finance, an open-data watchtower over Chile, and audio embeddings made visible.

ChronoScope, Univariate Forecasting Atlas with a Foundation-Model Tier
Scientific Machine Learning · Jul 2026

ChronoScope, Univariate Forecasting Atlas with a Foundation-Model Tier

A univariate time-series forecasting atlas: 15 diagnostic cases (4 real, licensed + 11 seeded synthetic), each forecast by the same 19-method ladder with backtested MASE/sMAPE/coverage, from classical baselines through ML and deep nets to four zero-shot foundation models (Chronos-Bolt, Chronos-2, TimesFM-2.5, TiRex-2). A 10-module statistical analysis suite runs alongside. Everything is baked offline and replayed in a static SPA; a Pyodide live lane forecasts a bring-your-own-series with a numpy core. The foundation models are offline-only, and on real M4-hourly a plain SeasonalNaive beats TimesFM-2.5.

scientific-ml time-series forecasting foundation-models +7
Atalaya, A Watchtower over Chile's Open Data
Data Engineering · Jun 2026

Atalaya, A Watchtower over Chile's Open Data

Atalaya harvests Chile's Data Observatory open catalog, profiles every downloadable table, and mines five kinds of cross-dataset relation (same-source, semantic similarity, spatial overlap, joinability and statistical correlation) into an explorable knowledge graph, with client-side semantic search running in the browser. Each edge carries its evidence strength, and none is presented as causal.

data-engineering knowledge-graph information-retrieval embeddings +2
Finn, Personal Finance & Market Forecasting
Quantitative Finance · Apr 2026

Finn, Personal Finance & Market Forecasting

A full-stack personal-finance and market-analytics app: track instruments, build watchlists and portfolios, run statistical and ML forecasts with walk-forward validation, review risk, and read FinBERT news sentiment, with LATAM coverage (IPSA + Chile macro) behind a fast, server-rendered UI.

quant forecasting finbert portfolio +4
Auralis, 6D Audio Embedding Visualizer
Audio & Signal · Apr 2026

Auralis, 6D Audio Embedding Visualizer

Turns any sound into a navigable 6-dimensional universe. Audio is mapped onto seven embedding tracks (interpretable spectral features, PCA/t-SNE/UMAP projections, the Tonnetz harmonic space, YAMNet event semantics, and CLAP audio-text meaning) and rendered as a luminous 3D trajectory you can fly through, across a curated library of 102 sounds.

audio embeddings visualization threejs +4

Haptics, Robotics & Accessibility

4

Haptic simulation and accessible interaction, and a robotic writer: research on how people and machines touch and move.

FeelIT 2.0, Haptic Accessibility Workbench
Accessibility & Haptics · Feb 2026

FeelIT 2.0, Haptic Accessibility Workbench

A web-based haptic accessibility workbench enabling visually impaired users to explore 3D objects through touch, read Braille text in immersive 3D environments, and navigate a haptic desktop. The modern realization of a vision begun in 2008.

accessibility haptics braille three-js +3
FeelIT, Haptic Accessibility Prototype (2008-2012)
Accessibility & Haptics · Jul 2012

FeelIT, Haptic Accessibility Prototype (2008-2012)

The original FeelIT project, a digital environment with tactile feedback for blind users to explore virtual objects through relief information. Pioneered a 10×10 electromagnetically controlled pin-array display and servo-motor haptic devices.

haptics accessibility virtual-reality hardware +1
3D Haptic Simulation with Octree Collision Detection
Accessibility & Haptics · May 2009

3D Haptic Simulation with Octree Collision Detection

A 3D simulation system recreating haptic interaction with virtual objects. Uses octree spatial partitioning (O(N log N)) and Separating Axis Theorem for collision detection with spring-damper Kelvin-Voigt force model.

haptics collision-detection octree three-js +3
Robotic Writer, 5-DOF Scorbot III Simulation
Robotics & Control · Dec 2008

Robotic Writer, 5-DOF Scorbot III Simulation

A full simulation and control environment for a 5-DOF Scorbot III robotic arm that picks letter blocks and spells words. Uses Denavit-Hartenberg convention with analytical closed-form inverse kinematics for real-time performance.

robotics kinematics denavit-hartenberg simulation +3

Education & Creative Instruments

4

Instruments built to teach and to make: a continuous botanical scene, a rigid-paper mechanism workshop, and virtual microscopy for science education.

Laminario, an Open Collection of Microscope Slides in the Spirit of iNaturalist
Education & Outreach · Sep 2026

Laminario, an Open Collection of Microscope Slides in the Spirit of iNaturalist

A public collection where every case is a real microscope slide shown as the glass object it is, its format and size, the specimen under the coverslip, a printed label with a scannable QR, and behind the glass the micro imagery, from one photomicrograph to a whole-slide scan with focal planes and polarised pairs, explored in deep zoom over IIIF. Three realms, 18 collections and 130 sub-collections and groups with their own iconography; visitors browse without an account, invited contributors add slides, the community agrees on identifications. Version 0.17.002: seventeen units released and deployed since 2026-09-30 at laminario.ml.fasl-work.com, with a base collection of 505 slides (786 images) from open sources, shown as realistic glass slides in 3D. Lifecycle planned.

microscopy slides natural-history iiif +6
Floraria, a Continuous Botanical Scene from Museum Scans to Cells
Education & Outreach · Sep 2026

Floraria, a Continuous Botanical Scene from Museum Scans to Cells

Point and scroll, or pinch, to open enclosing structures and approach tissues, cells and compartments of a flower in one continuous scene. Five Smithsonian CC0 bloom specimens at two delivery fidelities, two authored educational component models (a general flower and an orchid) with named structures, assembly and explosion, and twelve guided investigations that connect inspection, morphology, comparisons and the documented evidence, in English and Spanish, light and dark, with keyboard and touch controls, shareable selections and JSON and PNG export. The real-scan and authored-model lanes stay visibly labelled everywhere: museum scans show outer surfaces and establish no internal organ, no segmentation and no physical measurement. Lifecycle: building.

botany education 3d-visualization museum-data +3
Plega, a Rigid-Paper Mechanism Workshop from Editable Design to Actual-Size Cut Files
3D & Visualization · Sep 2026

Plega, a Rigid-Paper Mechanism Workshop from Editable Design to Actual-Size Cut Files

A paper-mechanism workshop that connects an original editable design to a moving rigid-paper view, explicit geometric diagnoses, pinned repair comparisons, actual-size vector export and assembly progress. Two restricted mechanism families, up to six separated lanes, six original starters and two deliberately invalid repair cases; every physical coordinate in millimetres; calibrated PDF, SVG, FOLD and JSON export; twelve editable cutwork compositions with transactional direct editing on the canvas. The model is rigid zero-thickness panels: it does not predict material stiffness, fold forces, glue performance or printer calibration, a lane-bound overlap means uncertified separation rather than a certain collision, and finite repair proposals do not claim global optimality. Lifecycle: building.

paper-engineering origami kinematics geometry +4
Education & Outreach · Dec 2014

Micromundo, Virtual Microscopy & Science Education Platform

A digital platform integrating education with virtual microscopy, making optical science accessible through low-cost 3D-printed microscopes ($50-100 vs $2,000+ professional) and web-based visualization tools. Backed by Start-Up Chile and Microsoft Imagine Lab.

education microscopy startup science-outreach +2

Mobile Apps

3

On-device applications: 3D reconstruction from a phone, and private mobile products.

Mobile Apps · Jul 2026

Croquis, On-Device 3D Reconstruction of Real Spaces on Android

An Android app that turns a walk-through (camera, IMU and GPS, no lidar) into a metric, consistency-gated point-cloud reconstruction computed fully on the phone and stored on the phone as a scene library you can view, measure and export, with a stated uncertainty budget. Imagery never leaves the device. Sensors own the trajectory and scale (ARCore VIO); the neural depth model contributes dense geometry, never the trajectory. In development (v0), Apache-2.0. The on-phone companion to the Lidar3D reconstruction line, with a local-GPU reprocessor (Croquis Station) above it.

mobile-apps android 3d-reconstruction on-device +6
Glissa, Expressive Glissando Instrument Proprietary
Mobile Apps · Jun 2026

Glissa, Expressive Glissando Instrument

A private mobile musical instrument whose signature is continuous glissando (slide smoothly between notes with Keys, Glide and Auto modes) backed by an FX rack, instrument morphing, sampled packs, and a Pro tier with cloud sync.

mobile audio music synthesis +3
Ronquy, On-Device Snore Detection Proprietary
Mobile Apps · Jun 2026

Ronquy, On-Device Snore Detection

A private mobile app that detects snoring on-device, overnight, with no audio leaving the phone, a YAMNet + TFLite model runs an 8-hour audio-and-inference loop locally, with an optional cloud account for sync.

mobile audio-ml on-device yamnet +3

Industry (proprietary)

7

Delivered for industrial clients under confidentiality: general description and a schematic only, no client data.

Hyperspectral Mineral Classification Platform Proprietary
Spectral Analysis · Oct 2025

Hyperspectral Mineral Classification Platform

A machine learning platform for classifying minerals and estimating abundances from hyperspectral imagery (VNIR/SWIR). Compresses turnaround from days to minutes for routine mineralogical characterization.

hyperspectral machine-learning mineral-classification cnn +2
Crusher Liner Wear Management System Proprietary
3D & Visualization · Aug 2025

Crusher Liner Wear Management System

A full-stack platform for tracking and forecasting crusher liner wear from raw 3D laser scan data through to production deployment. Replaced manual caliper measurements with automated 3D point cloud analysis.

point-cloud 3d-analysis wear-forecasting fastapi +2
Geotechnical Risk Prediction System Proprietary
Geotechnical & Risk · May 2023

Geotechnical Risk Prediction System

A machine learning system for predicting geotechnical hazards, rockburst and slope instability, in underground and open-pit mining. Provides weekly systematic risk assessment with SHAP-based explainability.

machine-learning geotechnical mining seismic +2
Environmental Monitoring & Mitigation System Proprietary
Computer Vision & AI · Dec 2022

Environmental Monitoring & Mitigation System

An environmental monitoring system combining computer vision, predictive modeling, and Generative AI to detect, forecast, and mitigate pollution events at mining operation sites. Achieved 15% reduction in severe alert events.

computer-vision generative-ai environmental monitoring +1
Real-Time Material Tracking & Blending System Proprietary
Mining & Optimization · May 2022

Real-Time Material Tracking & Blending System

A real-time material tracking and blending optimization system maintaining a digital twin of ore flow from mine pits through conveyors and stockpiles to processing plant inputs. Significantly improved blending compliance.

digital-twin optimization mining pyspark +2
Mining Process Optimization Platform Proprietary
Mining & Optimization · Dec 2021

Mining Process Optimization Platform

An end-to-end ML platform for optimizing mineral processing (SAG milling, flotation, thickening) deployed across multiple mining divisions. Delivered +100 TPH throughput uplift and measurable copper recovery improvements.

machine-learning mining optimization kedro +2
HyClus Viz, Hyperspectral Clustering Visualization Proprietary
Spectral Analysis · Aug 2019

HyClus Viz, Hyperspectral Clustering Visualization

Deep autoencoders combined with t-SNE for transforming raw hyperspectral data into interpretable visualizations. Achieved 95-97% accuracy for grain size classification from mining comminution feeders.

hyperspectral deep-learning dimensionality-reduction t-sne +2