Click to edit Master text styles

Second level

Third level

Fourth level

Fifth level

‹#›

(FIRST BULLET) Design and Develop prototype R.E.S.C.U.M.E. Applications

§Incident Scene Pre-Arrival Staging Guidance for Emergency Responders (RESP-STG)

§Incident Scene Work Zone Alerts for Drivers and Workers (INC-ZONE)

§Emergency Communications and Evacuation (EVAC) (Prototype development activities not initiated)

(THIRD BULLET) Data provided to support routing decisions includes:

•Satellite imagery

•GIS data

•Current weather data

(THIRD BULLET)

Warnings are provided when a vehicle approaching or in the incident zone is operated outside of safe parameters for the conditions.

System prototype developed and demonstrated

June 17, 2014 at the Ohio State University Airport – Don Scott Field in Columbus, OH

November 13, 2014 at Maryland Police and Correctional Training Commission’s Driver Training Facility located in Sykesville, MD

•New mapping data includes:

•Rich OSM Base Layer with Hospital Location & Government Buildings Identified

•Satellite Imagery with Lane Level Zoom Functionality

(FIRST BULLET) Developed the connected vehicle applications, which reside on separate vehicles (responder and oncoming) but work together to identify potential threats and collisions and warn drivers and responders in time to take evasive action

(THIRD BULLET) Implemented lane level mapping and Global Positioning System (GPS) positioning accuracy system This was required to support responder and oncoming vehicle imminent crash warnings as well as lane closure alerts and warnings in oncoming vehicles.

(FIRST BULLET) Developed and integrated DSRC, Cellular, and Bluetooth communications in both oncoming vehicle and responder vehicle systems. This included using smart phones and DSRC radios for comprehensive communications and efficient installation and operation

(SECOND BULLET) Address human factors components if delivering information, alerts, and warnings to responders and approaching drivers during the multitude of distractions inherent in an incident.

The R.E.S.C.U.M.E. Impact Assessment work includes three tracks:

•For INC-ZONE and RESP-STG, the assessment includes an evaluation of the prototype that was demonstrated in Maryland on November 13, 2014 through interviews with test participants.

•In terms of impacts of INC-ZONE and RESP-STG, the IA team is using modeling and simulation with the US 101 San Mateo Corridor to compare baseline performance measures (i.e. without R.E.S.C.U.M.E.) to performance measures with R.E.S.C.U.M.E. (i.e. with the incorporation of the INC-ZONE & RESP-STG functionalities). The unit benefits (or impacts) observed from the San Mateo simulation (i.e. for one incident) would be extrapolated to a regional level, and in this case the region would consist of the state of Maryland, since the demonstration occurred there, and the stakeholders involved in the prototype development were from that region. We will talk in more details about the simulation analysis of INC-ZONE & RESP-STG in the coming slides.

•For EVAC, the impacts are assessed through modeling and simulation using the Greater New Orleans evacuation model, with the TRANSIMS software. New Orleans was chosen because it provided actual data from a major evacuation event (that is Hurricane Katrina) as a baseline. Therefore, the baseline scenario will be the Katrina scenario (without any EVAC functionalities), while the various R.E.S.C.U.M.E. scenarios will include the coding of the EVAC functionalities in order to ultimately compare the performance with EVAC to the performance without EVAC, and thus assess the value of the application and its potential.

•The R.E.S.C.U.M.E. Impact Assessment team is led by Booz Allen Hamilton (Team lead: Gustave Cordahi). Booz Allen is conducting the assessment of INC-ZONE & RESP-STG. While a team comprised of Booz Allen, AECOM, and Prof. Brian Wolshon of LSU is working on the EVAC assessment.

•The INC-ZONE and RESP-STG applications are being assessed primarily using microscopic traffic simulation.

•The Testbed that is being modeled for this purpose is the US-101 freeway in San Mateo, California

•Testbed details:

- The corridor is an 8.5 mile long freeway which forms a major connector between San Jose and San Francisco.

- The corridor presents a variety of traffic control features such as HOV lanes, Ramp Metering etc. and carries an AADT (Average Annual Daily Traffic) between 200,000 to 250,000 vehicles.

- Data required for baseline calibration including incident data is available and is being used in the Testbed, including PeMS data, California Highway Patrol logs etc.

•Simulation Conditions used:

- 5-hour PM peak simulation is used for model calibration.

- Weather conditions used: Dry and Rainy weather conditions.

- Incident types: Short incident (30 minutes) and Long incident (60 minutes)

- Levels of market-penetration of the applications: 10%, 25%, 50%, 75% and 100%

•Modeling Capabilities:

- The modeling of INC-ZONE and RESP-STG applications is being done in VISSIM 7 traffic simulation software.

- Functionalities used:

•Lane-alerts and warnings are modeled as commands to change the vehicle’s desired lanes.

•Speed-alerts and warnings are modeled as commands to change the vehicle’s desired speed distribution.

This table highlights the corresponding modeling strategy used for different functionalities of the INC-ZONE and RESP-STG applications in the simulation environment.

INC-ZONE:

-Using instantaneous vehicle-positions in the simulation network (including link and lane information), a threat determination algorithm determines which vehicles should be given ‘alerts’ and ‘warnings’.

-Once a vehicle is found to be in the incident-lane or in the nearest lane and is approaching the incident, vehicle commands are provided to change the vehicle’s desired lane and speed distributions in order to enhance safety and mobility. These commands are based on the MUTCD guidelines on work-zone safety distances from the incident.

-Responder alerts and warnings are the warnings provided to the on-scene crew to enhance their safety. Due to model limitations, these functionalities are not modeled. However, we expect to see an enhancement in their safety due to the two other functionalities. We will be using surrogate safety measures guidelines previously published by USDOT to analyze this metric. This will be a post-processing effort.

RESP-STG:

-En-route staging plans are modeled using vehicle commands to deploy and stage the emergency vehicles near an incident for the duration of the incident.

-Emergency responder status reporting will be modeled as a performance monitoring variable to determine the travel-time and deployment time of emergency vehicles under different conditions.

•Performance Measures:

- The team is looking at both direct performance measures as well as indirect performance measures from the simulation.

- The three categories of performance measures that are being assessed are: safety, mobility and environmental.

•Direct performance measures:

- These are collected directly using simulation measurements and include mobility and environmental measures such as average vehicle delay, average travel-speed of the corridor, throughput of incident zones, fuel-consumption and emission of vehicles. Some surrogate safety measures are also provided by the simulation.

•Indirect performance measures:

- The indirect performance measures mainly include surrogate safety measures determined after the simulation runs using individual vehicle records. Trajectory analysis using SSAM (Surrogate Safety Assessment Model) would be done to assess safety implications of the INC-ZONE and RESP-STG applications. Some other safety measures that would be analyzed include the number of lane-changes in the vicinity of incident zone and the speed differential between lanes near the incident zone. Improvement in the response vehicle travel-time is also an indirect performance measure.

•Performance Measurement:

- Baseline measures (without R.E.S.C.U.M.E.) would be compared to measures with R.E.S.C.U.M.E., specifically – INC-ZONE alone, RESP-STG alone, and INC-ZONE and RESP-STG together.

- Finally, a regional extrapolation of the results would be performed using statistics from the RITIS database for Maryland.

•The EVAC impact assessment is being conducted through a modeling analysis that aims to assess the potential impact of EVAC functionalities on the evacuation in the Greater New Orleans area.

•The analysis is conducted by AECOM as a subcontractor to Booz-Allen-Hamilton with the assistance of Dr. Brian Wolshon of LSU

•

•The total evacuee population was generated based on Census data.

•The selection percentage represents the net result of the EVAC penetration rate (equipped vehicles) and the compliance rate

•The team is using 15 minute link travel times for dynamic routing calculated from a 6 second mesoscopic simulation of the region over 48 hours

•

•The major evacuation freeways include I-10 W to Baton Rouge, US-51 N to Hammond and I-12 and I-59 N to Hattiesburg.

•The transit routes include feeder bus routes and evacuation bus routes that transport evacuees out of the region.

•The upper figure shows the geographical split of evacuation demand among major evacuation destinations.

•The lower figure shows the temporal distribution of the evacuation demand. Evacuees generally choose to leave town during the day.

•Communication gaps might be expected at various times during the evacuation or at various locations where messages cannot be received. This applications does not address this consideration.

•Scenario 1 serves as the comparison baseline (i.e. the Katrina scenario without EVAC)

•In Scenario 2, evacuees with EVAC equipment are allowed to make en-route route choice by switching away from congested routes upon receipt of information about traffic conditions.

•In Scenario 3, an accident is assumed to happen and EVAC-equipped evacuees receive information about the accident and may divert away from the incident route.

•In Scenario 4, , evacuees with EVAC equipment may choose their destinations according to the information about capacity of the hotels and prevailing traffic conditions before departure. It is assumed that all evacuees will secure hotel lodging outside of study area. The data for the hotel accommodation capacity along major interstate corridors are collected and visualized on next slide.

•In Scenario 5, evacuees with EVAC equipment can go to the fueling station EVAC recommends based on available fuel and distance.

•In Scenario 6, Transit-based evacuees with EVAC equipment will adjust their departure time from home to minimize the wait time

•Scenario 7 investigates the synergetic effect of the EVAC functionalities by combining the modeling methodologies for all previous scenarios

•

The circles represent the number of hotel rooms in each of the major towns or cities on evacuation routes outside of New Orleans

•