gui4gmns — AI-guided dashboards for GMNS

The process — decompose it forward, then transfer to any city

Forward process: OSM to GMNS network, then demand, assignment, integrate (GTFS/signals/sensors/probe), then gui4gmns generate to dashboard + figures + portals; transferable to any city

Five steps forward (osm2gmns → demand → assignment → integrate → visualize); swap the city's inputs and the steps + dashboards stay the same. ▶ see it run on Tempe, AZ · details in the GMNS → dashboards skill.

The output — one GMNS folder opens in every portal

One GMNS folder to many visualization portals

▶ Live deck.gl demo Chicago Sketch network, in your browser — no install Gallery Interactive dashboards + static figures Dashboard folder All self-contained dashboards, one place Portals guide Online/offline, 3D & trajectories, student how-to Source on GitHub Code, datasets, pip install gui4gmns Datasets What each demo network exercises Template gallery Catalog-driven dashboard templates by category Skill: GMNS → dashboards Convert any city (package + LLM), data elements & QA ▶ Integrated dashboard One map: switch city, basemap, MoE (volume/speed/V-C), OD lines, click-to-inspect ▶ AZ city dashboard Switch across 7 Phoenix-metro cities on one map (OSM) ▶ GMNS 3D (beta) volume→height, speed→color — traffic overlay for a 3D city base (strategy) ▶ I-405 N — observed (Caltrans PeMS) Real 5-min speed+flow, AM peak — recurrent bottleneck from measured detectors ▶ I-405 N — QVDF model Calibrated QVDF speed from inflow demand/capacity — full day, D/C ratio on hover ▶ I-95 3D TMC timeline Real 24h×15-min speeds — play the corridor: baseline → bottleneck → spillback → recovery ▶ Event playbook (queue + incidents) Time slider with measured queue spillback + blocked-link events firing on the clock ▶ I-880 N — incidents (Caltrans PeMS + CHP) Oakland AM peak — real 5-min speeds with CHP incidents firing on the timeline: baseline → collision → spillback ▶ TrafficFlowBench Visual Analytics Inspect network, state, queues, OD/path flows & incident evidence from GMNS freeway benchmark data TMC playbook Operations vocabulary, event storyline, visual grammar & fidelity ladder

See it — real networks in real portals

ITS I-95 VA corridor colored by observed INRIX speed — **ITS I-95 (VA) data hub** ★ — the flagship: network + INRIX speed + VDOT sensors + probe, colored by observed speed (ramps/interchanges slow, red).
▶ full data hub · Kepler.gl · deck.gl · Google Earth

Tempe AZ network colored by assignment volume — **Tempe, AZ** — built end-to-end from OSM: `osm2gmns → demand → assignment`, colored by assigned volume (arterials red, local streets green). The transferable process, run on a real city.
▶ Kepler.gl · deck.gl · Google Earth · files

Chicago Sketch network colored by volume — **Chicago Sketch** — 2,950 links, colored by volume.
▶ Kepler.gl · deck.gl · Google Earth (KML) · files

ARC Atlanta regional network colored by volume — **ARC Atlanta regional** — 145,971 links; the I-285 loop + I-75/85/20 spokes load red. (Live map = top 6,000 by volume, for speed.)
▶ Kepler.gl · deck.gl · Google Earth (KML) · files

Arizona — PHX metro, city by city (OpenStreetMap only)

Every AZ network here is built only from public OpenStreetMap via osm2gmns (each fetched from OSM by city, clipped to its real OSM city boundary — blue outline). No agency travel model. ▶ open the city-by-city dashboard — one page, switch across all seven cities on an interactive map. Every map is stamped "Source: OpenStreetMap via osm2gmns."

Phoenix AZ street network from OpenStreetMap, clipped to the city boundary — **Phoenix, AZ** — the full city street network from OpenStreetMap (158,570 links, clipped to the OSM city boundary), top 6,000 by free-flow speed.
▶ Kepler.gl · deck.gl · Google Earth

osm2gmns raw big map, all OSM streets, Tempe AZ — **osm2gmns "big map" — step 1** — the raw network straight from OpenStreetMap: every street (freeways green, arterials yellow, dense residential red). The decompose step on its own, 22,296 links.
▶ Kepler.gl · deck.gl · Google Earth

Mesa AZ network from OSM — **Mesa** — ▶ Kepler · deck.gl · KML

Scottsdale AZ network from OSM — **Scottsdale** — ▶ Kepler · deck.gl · KML

Chandler AZ network from OSM — **Chandler** — ▶ Kepler · deck.gl · KML

Glendale AZ network from OSM — **Glendale** — ▶ Kepler · deck.gl · KML

Gilbert AZ network from OSM — **Gilbert** — ▶ Kepler · deck.gl · KML

Offline by default (open the dashboard with no internet); online portals — Kepler.gl, deck.gl, Google Earth — add richer 3D and trajectory exploration. Everything here is generated by AI from a GMNS folder; ask it for the same views on your own data.