Optimizing people flow and safety in buildings Harri

Optimizing people flow and safety in buildings Harri Ehtamo, Aalto University, Espoo, Finland harri. ehtamo@aalto. fi Co-authors: Juha-Matti Kuusinen, Janne Sorsa, Marja-Liisa Siikonen, Henri Hakonen KONE Corporation INFORMS Annual Meeting 2018, Nov. 4 -7, 2018 The document can be stored and made available to the public on the open internet pages of Aalto University. All other rights are reserved.

The aim of the talk To describe key advances in people flow and building safety modeling. Based on several years of collaborative research carried out by: KONE Corporation and Systems Analysis Laboratory of the Aalto University. 2

ORMS – Today: 44/2, April 2017 People flow in buildings 3

Elevators form the core of the buildings Nowadays, 55 percent of the total global population is living in an urban environment; in 2050, 68 percent. Elevators form the core of the buildings. Their value largely depends on their performance; and on the vertical transportation system as a whole. The quality of a building depends on the equipment for smooth and efficient people flow; and on safety: e. g. , fast evacuation in an emergency situation. 4

Background research Optimization models and numerical algorithms for an Elevator Group Control System (EGCS). For example: Optimal control of double-deck elevator group using genetic algorithms, Sorsa et al. (2003); Modeling uncertain passenger arrivals in the elevator dispatching problem with destination control, Sorsa et al. (2017). Since 2003: 5 M. Sc. theses, 3 D. Sci. (Tech. ) theses. 5

Passenger traffic forecasts EGCS measures the people flow by counting the number of boarding and alighting passengers. The floor- and direction based statistics: passenger number measurements for each floor and direction throughout the day. EGCS forms the traffic forecasts from the measurements and uses them to dispatch the elevators to the calls in an efficient way. Traffic forecasts: The elevator trip origin-destination (OD) matrix estimation problem, Kuusinen et al. (2015). 6

An elevator trip with five nodes and five OD pairs. plus the flow conservation constraints: 7

A building OD matrix Mixed lunch hour traffic in a 25 -floor office building; a 15 min simulation. 8

Sociality completes traffic planning In a real world passengers arrive, during an up-peak traffic, according to a Poisson process, Alexandris (1972). Kuusinen et al. (2012) show that this assumption does not hold throughout the day, e. g. , during the mixed lunch time traffic. Rather, in high-rise buildings passengers follow a compound Poisson process for batches, with well defined batch-size distribution depending on the time of day. 9

Morning and lunch time batches The batch-size distribution depends on the time of day. 10

KONE BTS, KONE Building Traffic Simulator 11

Building safety and evacuation research Important question evacuees face: what exit to use. In agent-based evacuation simulation models, the exit selection is based on a rule, or a more advanced algorithm. The agents may choose a familiar exit, or observe the situation and select the nearest exit. Our experiments, Heliövaara et al. (2012) suggest that people in an evacuation may not be able to choose the fastest exit, nor do they behave altruistically. 12

They may not be able to behave altruistically A trial of the experimental study in evacuation of a corridor. 13

Fire Dynamics Simulator with Evacuation, FDS+Evac is able to create realistic behavior in a crossing. 14

Why is the back row rushing? Stable equilibria curves in a spatial game of patient (yellow), and impatient (black) agents under threatening conditions. 15

Related literature Alexandris, N. A. , 1977, “Statistical models in lift systems”, Ph. D. thesis, University of Manchester, Institute of Science and Technology. Helbing, D. , Farkas, I. , Vicsek, T. , 2000, “Simulating dynamical features of escape panic”, Nature, Vol. 407, pp. 487 -490. Heliövaara, S. , Korhonen, T. , Hostikka, S. , Ehtamo, H. , 2012 a, “Counterflow model for agent-based simulation of crowd dynamics”, Building and Environment, Vol. 48, pp. 89 -100. Heliövaara, S. , Kuusinen, J. -M. , Rinne, T. , Korhonen, H. , Ehtamo, H. , 2012 b, “Pedestrian behavior and exit selection in a corridor – An experimental study”, Safety Science, Vol. 50, pp. 221 -227. Heliövaara, S. , Ehtamo, H. , Helbing, D. , Korhonen, T. , 2013, “Patient and impatient pedestrians in a spatial game for egress congestion”, Physical Review E, Vol. 87, pp. 012802. Korhonen, T. , Hostikka, S. , 2009, “Fire dynamics simulator with evacuation: FDS+Evac – Technical reference and user’s guide”, VTT working papers 119, VTT Technical Research Centre of Finland. 16

Kuusinen, J. -M. , Sorsa, J. , Siikonen, M. -L. , 2015, “The elevator trip origin-destination matrix estimation problem”, Transportation Science, Vol. 49, pp. 559 -576. von Schantz, A. , Ehtamo, H. , Pärnänen, I. , 2017, “Twotype multiagent game for egress congestion”, Hawaii International Conference on System Sciences (HICSS-50). von Schantz, A. , Ehtamo, H. , “Overtaking others in a spatial game of exit congestion”, submitted to Scientific Reports. Siikonen, M. -L. , 1997, “Planning and Control Models for Elevators in High-Rise Buildings”, Ph. D. thesis, Helsinki University of Technology, Systems Analysis Laboratory. Sorsa, J. , Siikonen, M. -L. , Ehtamo, H. , 2003, “Optimal control of double-deck elevator group using genetic algorithm”, International Transactions in Operational Research, Vol. 10, pp. 103 -114. Sorsa, J. , Ehtamo, H. , Kuusinen, J. -M. , Ruokokoski, M. , Siikonen, M. -L. , 2017, “Modeling uncertain passenger arrivals in the elevator dispatching problem with destination control”, Optimization Letters, Vol. 12, pp. 171 -185. Tyni, T. , Ylinen, J. , 2001, “Genetic algorithms in elevator car routing problem”, in: Spector, L. Goodman, E. D. , Wu, A. , Langdon, W. B. , Voigt, H. -M. , Gen, M. , Sen, S. , Dorigo, M. , Pezeshk, S. 17