A case examine – Google Analysis Weblog

Some cities or communities develop an evacuation plan for use in case of an emergency. There are a selection of the reason why metropolis officers may enact their plan, a main one being a pure catastrophe, similar to a twister, flood, or wildfire. An evacuation plan may also help the neighborhood extra successfully reply to an emergency, and so might assist save lives. Nevertheless, it may be tough for a metropolis to guage such a plan as a result of it’s not sensible to have a whole city or metropolis rehearse a full blown evacuation. For instance, Mill Valley, a metropolis in northern California, created a wildfire evacuation plan however lacked an estimate for the way lengthy the evacuation would take.

At this time we describe a case examine through which we teamed up with the town of Mill Valley to check and enhance their evacuation plan. We define our strategy in our paper, “Mill Valley Evacuation Study”. We began by utilizing a site visitors simulator to mannequin a citywide evacuation. The analysis objective was to offer the town with detailed estimates for the way lengthy it will take to evacuate the town, and, by learning the egress sample, to seek out modifications to make the plan simpler. Whereas our prior work on this topic supplied an estimate for the evacuation time and confirmed how the time could possibly be decreased if sure highway modifications have been carried out, it seems the suggestions in that paper — similar to altering the variety of outgoing lanes on an arterial — weren’t possible. The present spherical of analysis improves upon the preliminary examine by extra precisely modeling the quantity and beginning areas of automobiles, by utilizing a extra practical map, and by working intently with metropolis officers to make sure that really helpful modifications to the plan are deemed viable.

Geography and methodology

Mill Valley is in Marin County, California, north of San Francisco. Lots of the residences are situated on the steep hillsides of a number of valleys surrounded by dense redwood forests.

Aerial views of Mill Valley, courtesy of the Metropolis of Mill Valley.

Lots of these residences are in areas which have just one exit path, towards the city heart. From there the perfect evacuation route is towards Freeway 101, which is within the flat a part of the town and is the most probably space to be removed from potential wildfires. Some neighborhoods produce other routes that lead away from each the town and Freeway 101, however these routes move via hilly forested areas, which could possibly be harmful or impassable throughout a wildfire. So, the evacuation plan directs all automobiles west of Freeway 101 to go east, to the freeway (see map under). The neighborhoods east of Freeway 101 should not included within the simulation as a result of they’re away from areas with a excessive hearth hazard score, and are near the freeway.

Mill Valley has about 11,400 households west of Freeway 101. Most Mill Valley households have two automobiles. Evacuation instances scale with the variety of automobiles, so it’s within the widespread curiosity to attenuate the variety of automobiles used throughout an evacuation. To that finish, Mill Valley has a public consciousness marketing campaign geared toward having every family evacuate in a single automobile. Whereas nobody is aware of what number of automobiles could be used throughout an evacuation, it’s protected to imagine it’s on common between one and two per family. The fundamental evacuation drawback, then, is easy methods to effectively get between 11 and 23 thousand automobiles from the assorted residences onto one of many three units of Freeway 101 on-ramps.

The simulated a part of Mill Valley west of Freeway 101 is contained in the blue border. Freeway 101 is proven in inexperienced. The crimson squares point out the three units of Freeway 101 on-ramps. The pink space has the best hearth hazard score.

The present work makes use of the identical common methodology because the earlier analysis, specifically, operating the open source SUMO agent-based traffic simulator on a map of Mill Valley. The site visitors simulator fashions site visitors by simulating every automobile individually. The detailed behaviors of automobiles are dictated by a car-following model. Every automobile is given some extent and time at which to begin and an preliminary route. The routes of most automobiles are up to date all through the simulation, relying on situations. To contemplate potential modifications in driver conduct below the excessive stress situations of an evacuation, the consequences of the “aggressiveness” of every automobile can also be investigated, however in our case the impacts are minimal. Some simplifying assumptions are that automobiles originate at residential addresses and the roads and highways are initially empty. These assumptions correspond roughly to situations that could possibly be encountered if an evacuation occurs in the course of the night time. The principle inputs within the simulation are the highway community, the family areas, the common variety of automobiles per family, and a departure temporal distribution. We have now to make assumptions in regards to the departure distribution. After discussing with the town officers, we selected a distribution such that almost all automobiles depart inside an hour.

4 bottlenecks

Mill Valley has three units of Freeway 101 on-ramps: northern, center, and southern. All of the automobiles should use considered one of these units of on-ramps to achieve their vacation spot (both the northernmost or southernmost phase of Freeway 101 included in our map). Provided that we’re solely involved with the vast majority of Mill Valley that lies west of the freeway, there are two lanes that strategy the northern on-ramps, and one lane that approaches every of the center and southern on-ramps. Since each automobile has to move over considered one of these 4 lanes to achieve the freeway, they’re the bottlenecks. Given the geography and present infrastructure, including extra lanes is infeasible. The intention of this analysis, then, is to attempt to modify site visitors patterns to maximise the speed of site visitors on every of the 4 lanes.

Evacuation plan

After we began this analysis, Mill Valley had a preliminary evacuation plan. It included modifying site visitors patterns — disabling site visitors lights and altering site visitors guidelines — on a couple of highway segments, in addition to specifying the sources (site visitors officers, signage) essential to implement the modifications. For instance, a two-way highway could also be modified to a one-way highway to double the variety of outgoing lanes. Quickly altering the path of site visitors is known as contraflow.

The plot under reveals the simulated fraction of automobiles which have departed or reached their locations versus time, for 1, 1.5, and a pair of automobiles per family (left to proper). The dashed line on the far left reveals the fraction which have departed. The stable black traces present the preliminary evacuation plan outcomes and the dotted traces point out the traditional highway community (baseline) outcomes. The preliminary evacuation plan considerably accelerates the evacuation.

The cumulative fraction of automobiles vs. time in hours. The demand curve is proven within the dashed line on the far left. The stable traces present the preliminary evacuation plan curves for 1, 1.5 and a pair of automobiles per family (left to proper). The dotted traces present the identical for the baseline case.

We are able to perceive how efficient the preliminary evacuation plan is by measuring the charges on the bottlenecks. The under plots present the speed of site visitors on every of the 4 lanes resulting in the freeway on-ramps for the case of 1.5 automobiles per family for each the baseline case (the traditional highway guidelines; proven shaded in grey) and the preliminary evacuation plan (proven outlined in black). The common fee per lane varies tremendously within the totally different instances. It’s clear that, whereas the evacuation plan results in elevated evacuation charges, there’s room for enchancment. Particularly, the center on-ramps are fairly underutilized.

The charges of site visitors on the 4 lanes resulting in Freeway 101 on-ramps for each the baseline case (regular highway guidelines; proven shaded in grey) and the preliminary evacuation plan (proven outlined in black).

Closing evacuation plan

After learning the map and investigating totally different alternate options, we, working along with metropolis officers, discovered a minimal set of recent highway modifications that considerably decrease the evacuation time in comparison with the preliminary evacuation plan (proven under). We name this the ultimate evacuation plan. It extends the contraflow part of the preliminary plan 1000 ft additional west, to a predominant intersection. Crucially, this permits for one of many (usually) two outgoing lanes to be devoted to routing site visitors to the center on-ramps. It additionally creates two outgoing lanes from that predominant intersection clear via to the northern on-ramps, over ¾ of a mile to the east.

A map of the principle modifications within the last evacuation plan. The crimson line reveals that site visitors heading north on Camino Alto will get diverted to the center Freeway 101 on-ramps. The blue line reveals site visitors within the northern lane of E Blithedale Ave will get routed on the brand new contraflow part.

The speed per lane plots evaluating the preliminary and last evacuation plans are proven under for 1.5 automobiles per family. The simulation signifies that the ultimate plan will increase the common fee of site visitors on the lane resulting in the center on-ramps from about 4 automobiles per minute to about 18. It additionally will increase the via fee of the northern on-ramps by over 60%.

The charges of site visitors on the 4 lanes resulting in Freeway 101 on-ramps for each the preliminary case (proven shaded in grey) and the ultimate evacuation plan (proven outlined in black).

The under plot reveals the cumulative fraction of automobiles vs. time, evaluating the instances of 1, 1.5 and a pair of automobiles per family for the preliminary and last evacuation plans. The speedup is sort of vital, on the size of hours. For instance, with 1.5 automobiles per family, it took 5.3 hours to evacuate the town utilizing the preliminary evacuation plan, and solely 3.5 hours utilizing the ultimate plan.

The cumulative fraction of automobiles vs. time in hours. The demand curve is proven within the dashed line on the far left. The stable traces present the ultimate evacuation plan curves for 1, 1.5 and a pair of automobiles per family (left to proper). The dotted traces present the identical for the preliminary evacuation plan.

Conclusion

Evacuation plans could be essential in shortly getting many individuals to security in emergency conditions. Whereas some cities have site visitors evacuation plans in place, it may be tough for officers to find out how effectively the plan works or whether or not it may be improved. Google Analysis helped Mill Valley check and consider their evacuation plan by operating site visitors simulations. We discovered that, whereas the preliminary plan did velocity up the evacuation time, some minor modifications to the plan considerably expedited evacuation. We labored intently with the town throughout this analysis, and Mill Valley has adopted the ultimate plan. We have been capable of present the town with extra simulation particulars, together with outcomes for evacuating the town one space at a time. Full particulars could be present in the paper.

Detailed suggestions for a selected evacuation plan are essentially particular to the world below examine. So, the particular highway community modifications we discovered for Mill Valley should not immediately relevant for different cities. Nevertheless, we used solely public information (highway community from OpenStreetMap; family data from census information) and an open supply simulator (SUMO), so any metropolis or company might use the methodology utilized in our paper to acquire outcomes for his or her space.

Acknowledgements

We thank former Mayor John McCauley and Metropolis of Mill Valley personnel Tom Welch, Lindsay Haynes, Danielle Staude, Rick Navarro and Alan Piombo for quite a few discussions and suggestions, and Carla Bromberg for program administration.