Dataset connection

Checking local data service…

The webpage is hosted online. Trajectories are read by a local Python service from your external drive; they are not uploaded to this website.

100AV EXPERIMENT / 01

See the traffic. Find the event.

Explore trajectories across time and space, then inspect the vehicles behind them.

DATA EXPLORATION
All times: Nashville · America/Chicago
Reading dataset inventory…
Connecting to local dataset…
—
—
—
—

Space–time trajectories

Click a line to inspect a vehicle · drag across time to select a window, then load

Camera-covered road section

I-24 MOTION camera map · teal rectangle locates the section shown in the trajectory plot

Map: I-24 MOTION v1.x documentation, Fig. 1 · Source

Your location within the camera corridor

Blue = approximate main camera corridor (MM 58.8–62.8). Teal rectangle = selected section. The separate validation poles farther southeast are not included.

Loading corridor context…

Drag horizontally on the overview to select a section, then zoom. Map rectangle is approximate, using x / 5280 to estimate mile marker under the documented v1.x convention, then interpolating the published map markers. It is a road-section locator, not a calibrated camera field-of-view polygon or a GPS transformation. Coverage varies by data/run; the plot may include positions beyond the map’s main camera markers.

Low → high speed · native x units/s
Hover over a trajectory to see its measurements.

Road playback

Native road coordinates · vehicle points, without inferred lane boundaries

—

Vehicle inspector

Choose a vehicle in the space–time plot.

Speed across time and location

Mean of displayed valid speed samples in each bin · empty bins have no observations

AV-centered traffic inspection

Choose one experiment vehicle to inspect its track and nearby traffic in a shared time and location window

CHOOSE AN AV

Choose an experiment vehicle

One selector for both the traffic neighborhood and geographic track. Grey IDs 7, 49 and 81 have no onboard GPS files in this folder; this does not establish absence from the experiment. Choose a numbered vehicle to load its best nearby usable passage. Additional original source IDs are preserved below.

Reference time plus look-back/look-forward sets the time window. The automatic neighborhood follows the projected AV at each time, retaining traffic within the upstream/downstream margins. It excludes intervals without supported GPS interpolation (gaps over 3 seconds). A custom mile-marker interval sets a fixed road section; mile markers are approximate.

Choose a numbered car to inspect its candidate neighborhood.

GPS and alignment details

This is a candidate neighborhood for inspection. The fleet track is projected with the official v1-2 median spline; its compatibility, GPS time offset, reference-point differences and AV↔MOTION identity are not yet verified. Nearby records are not confirmed human drivers, following vehicles, or affected vehicles. The center line marks an inspection time, not a verified controller switch. Before/after here means relative to that time, not a causal treatment comparison. Source event time and any automatic jump to a passage are shown explicitly.

Geographic GPS track of the same selected AV

Choose a numbered AV above. This view uses the same vehicle and loaded time window.

Click a GPS point to inspect its location.