Test 5: Assigned occupant demographics

1. Definition

A square room of 100 m by 100 m holds 100 occupants spread evenly. Walking speeds are assigned from a population chosen to match the expected building occupancy, and the data source is documented (ISO points to Lord et al., 2005, for example distributions). The test checks that the walking speeds the model produces agree with the assigned distribution, over several runs, with a statistical method chosen by the tester.

Source: ISO 20414:2020, Test 5, Table 6, pp. 14-15. This is a paraphrase, not the text of the standard.

Reference: ISO (2020). ISO 20414:2020, Fire safety engineering - Verification and validation protocol for building fire evacuation models. First edition. International Organization for Standardization, Geneva. https://www.iso.org/standard/78348.html (Page numbers are the printed page numbers of ISO 20414:2020 (the copy read is a proof, so numbers can differ by a page in the published edition). ISO states that its tests are examples and that a tester may use different tests that verify the same component equivalently.)

2. Scenario

Scenario figure from the executed notebook.

3. What to expect

The simulated walking-speed distribution is consistent with the distribution specified in the scenario.

Our criterion: Realized speeds match Gaussian(1.2, 0.2) (pooled mean/std + KS)

4. Our result

Status: PASS

Test cases of this row

• test_represented_speed_matches_gaussian: bound: 1.2 - 0.05 < mean_speed_mps < 1.2 + 0.05; bound: 0.2 - 0.05 < std_speed_mps < 0.2 + 0.05; bound: median_ks_p > 0.05 (pooled over 3 seeds)

Result figure from the executed notebook.

5. Our implementation

Deviations from the standard:

• Desired speeds follow a Gaussian with mean 1.2 m/s and standard deviation 0.2 m/s. This is the suite’s choice and is not tied to a published building population.
• The scenario file has no exit. The test adds a full-width exit along the x = 100 m wall and measures each agent’s median step speed while it walks the room.
• The pass rule is the suite’s own: over 3 seeded runs, the pooled mean and standard deviation of the measured speeds are each within 0.05 m/s of the assigned values, and the median of the per-run two-sample Kolmogorov-Smirnov p-values is above 0.05.
• Agents are placed at random positions in the central 90 m by 90 m area, not on a regular grid.

Model: CollisionFreeSpeedModel

• Test function: tests/vv/test_iso.py::test_represented_speed_matches_gaussian
• Scenario file: Iso-05-demographics.zip

6. Our notebook

• Executed notebook (HTML)
• Notebook source (.ipynb) on GitHub