Exit choice and familiarity

Which exit an occupant heads for is not decided by distance alone. The studies below find that occupants tend to leave the way they know, are influenced by what other people do, and respond to smoke, lighting and crowding at the exits. There is no single published law for exit choice; there are observations and calibrated statistical models.

Movement to the familiar

Sime (1985) studied the direction of escape in a fire in a large room with an entrance and an emergency exit in opposite corners. He contrasted the “affiliative” model, in which people under threat of entrapment move towards familiar persons and places, with design assumptions that the physical availability and proximity of an exit determine its use. Proximity mattered, but so did affiliation: staff generally left by the fire exit, whereas members of the public who were separated from their group moved towards, and left by, the entrance. Sime argued that place affiliation is not addressed sufficiently in escape-route design.

Edelman, Herz and Bickman (1980, pp. 193–196) interviewed 22 residents after a nursing-home fire. Up to about 85 residents (93 % of the floor) left by the one stair that staff and residents used every day, most of them towards the fire. Yet most respondents knew the exits at the end of their own wing, and 5 of the 6 in the fire zone named a closer one. Those exits were alarmed and labelled “Emergency Exit Only”, and 13 of 14 respondents had never used them. Nine of 13 said they moved because staff told them to leave, probably without saying which exit to use, and 11 of 18 saw other residents going the same way. The authors found no lack of awareness of the exits; the residents had no practice using them (p. 195).

Kinateder, Comunale and Warren (2018) tested this in an ambulatory virtual museum. Participants entered through one door and, when an alarm sounded, were significantly more likely to leave through that familiar door than through a second exit. The effect grew when virtual neighbours also left by the familiar door, shrank when they left by the other door, and the social influence was stronger with two neighbours than with one.

Discrete-choice models

Lovreglio, Borri, dell’Olio and Ibeas (2014) introduced a random-utility discrete-choice model for exit choice in emergency evacuations. Haghani and Sarvi (2017, p. 241) describe it as an internet-based stated-choice experiment with binary choices in simple visualised geometries, which investigates the effect of individuals’ demographic characteristics on exit choice. Lovreglio, Fonzone and dell’Olio (2016) calibrated a mixed logit model on an online stated-preference survey with non-immersive virtual reality and 1503 participants. Smoke, emergency lighting, exit distance, the number of evacuees near the exits and near the decision-maker, and the flow of evacuees through the exits all affected local exit choice significantly, with a high degree of behavioural uncertainty.

Haghani and Sarvi (2017) compared stated choices (4958 observations from face-to-face interviews in three public places) with revealed choices (3015 exit choices extracted from video of evacuation trials in which participants competed in a real crowd). The four data sets gave fairly similar patterns of parameter estimates, and the stated-choice models predicted choices reasonably similar to the revealed-choice model, despite significant differences in parameter scale. Their review of empirical methods in crowd research (Haghani and Sarvi 2018) surveys more than 160 studies.

Known limits

Apart from incident studies such as Sime’s and Edelman et al.’s, the evidence above comes from hypothetical choices, virtual reality and evacuation trials, none of which carries the threat of a real fire. Whether parameters calibrated in one geometry and population transfer to another is an open question; Haghani and Sarvi (2017) set out to test exactly this context-dependence. The guide to FDS+Evac, the evacuation module of the Fire Dynamics Simulator (FDS), notes, citing the socio-psychological literature, that familiarity of exit routes is an essential factor in evacuees’ decisions and that emergency exits are rarely used in many real evacuations because they are unfamiliar (Korhonen 2021, §3.5).

Sources

• Sime, J. D. (1985). Movement toward the familiar: person and place affiliation in a fire entrapment setting. Environment and Behavior, 17(6), 697–724. doi:10.1177/0013916585176003
• Edelman, P., Herz, E., & Bickman, L. (1980). A model of behaviour in fires applied to a nursing home fire. In D. Canter (Ed.), Fires and Human Behaviour (pp. 181–203). John Wiley & Sons, Chichester. ISBN 0-471-27709-6. No DOI or public URL.
• Kinateder, M., Comunale, B., & Warren, W. H. (2018). Exit choice in an emergency evacuation scenario is influenced by exit familiarity and neighbor behavior. Safety Science, 106, 170–175. doi:10.1016/j.ssci.2018.03.015
• Lovreglio, R., Borri, D., dell’Olio, L., & Ibeas, A. (2014). A discrete choice model based on random utilities for exit choice in emergency evacuations. Safety Science, 62, 418–426. doi:10.1016/j.ssci.2013.10.004
• Lovreglio, R., Fonzone, A., & dell’Olio, L. (2016). A mixed logit model for predicting exit choice during building evacuations. Transportation Research Part A: Policy and Practice, 92, 59–75. doi:10.1016/j.tra.2016.06.018
• Haghani, M., & Sarvi, M. (2017). Stated and revealed exit choices of pedestrian crowd evacuees. Transportation Research Part B: Methodological, 95, 238–259. doi:10.1016/j.trb.2016.10.019
• Haghani, M., & Sarvi, M. (2018). Crowd behaviour and motion: empirical methods. Transportation Research Part B: Methodological, 107, 253–294. doi:10.1016/j.trb.2017.06.017
• Korhonen, T. (2021). Fire Dynamics Simulator with Evacuation: FDS+Evac. Technical Reference and User’s Guide (FDS 6.7.6, Evac 2.6.0 draft), §3.5. VTT Technical Research Centre of Finland. github.com/tkorhon1/FDS-Evac-Guide. Secondary source.

How pyFDS-Evac uses this: see route rerouting and wayfinding.

How it is verified: Familiarity and the S4 T-junction test.