Proposed Watertown Plan Road Interchange Evaluation Using Full Scale Driving Simulator

Transcription

1 Proposed Watertown Plan Road Interchange Evaluation Using Full Scale Driving Simulator Kelvin R. Santiago-Chaparro*, M.S., P.E. Assistant Researcher Traffic Operations and Safety (TOPS) Laboratory Department of Civil and Environmental Engineering University of Wisconsin-Madison Engineering Drive B Madison, WI 0 ksantiago@wisc.edu Dan Reichl, M.S. Research Assistant Traffic Operations and Safety (TOPS) Laboratory Engineering Drive B Madison, WI 0 reichl@wisc.edu Andrea Bill, M.S. Associate Researcher Traffic Operations and Safety (TOPS) Laboratory Engineering Drive B Madison, WI 0 bill@wisc.edu David A. Noyce, Ph.D, P.E. Professor Traffic Operations and Safety (TOPS) Laboratory Department of Civil and Environmental Engineering Engineering Drive 0 Madison, WI 0 noyce@engr.wisc.edu *Corresponding Author Word Count:,0 words in text + ( figures * 0) =,

2 Santiago-Chaparro, Reichl, Bill, and Noyce 0 ABSTRACT With the increasing popularity of driving simulators, an opportunity has arisen to introduce the technology into the road design process. By exposing subjects to a proposed/existing design in a driving simulator a better understanding of how a design influences the performance of road users can be obtained without the need for the project to be operational. Furthermore, the preconstruction evaluation adds a behavioral component to the design process allowing for the construction of safer roads. Additionally, by identifying design issues during the design process expensive changes in a post-construction stage can be reduced/avoided. The aforementioned concepts were applied to the proposed Watertown Plank Road interchange project in Wisconsin. The Watertown Plank Road and US Highway interchange is part of the Zoo interchange project in Wisconsin. Using the design drawings provided by the design team researchers converted the drawings into a three dimensional model capable of running on a full scale driving simulator and used it to study driving behavior on the proposed project before the design process was finished. By combining driving behavior data collected using the full scale driving simulator and eye tracking data researchers were able to identify design concerns in the proposed project. The identification of design concerns prior to construction allowed the design team to make changes to the design prior to the start of the project thus avoiding changes later in the construction process.

3 Santiago-Chaparro, Reichl, Bill, and Noyce 0 INTRODUCTION The University of Wisconsin - Madison Traffic Operations and Safety (TOPS) Laboratory was selected by the Wisconsin Department of Transportation (WisDOT) to evaluate the proposed design of the US Highway (USH) and Watertown Plank Road interchange through the use of a full-scale driving simulation study. Using the proposed highway design drawings the research team created a virtual three dimensional environment compatible with driving simulators relying on the Realtime Technology (RTI) system. The scenario created by the research team provided subjects with the opportunity to be immersed in a driving environment that mimics the behavior of a real vehicle on a proposed highway. Figure depicts the driving simulator used by the research team to conduct the experiment. The driving simulator is mounted on a one degree of freedom platform that provides subjects with the feeling of acceleration and deceleration. Furthermore, because of the cylindrical screen, the realism of the simulation is increased since no screen edges are visible as in the case of simulators that use multiple projection screens to display the scenario. Specifications: 0 Cylindrical Screen & feet tall Instrumented Ford Fusion Body Eye Tracking Equipment 0 0 Figure. Full Scale Driving Simulator The evaluation process undertaken by the research team, design team, and WisDOT was the first one of its type attempted by WisDOT. Therefore, in addition to the evaluation of the proposed design the research team had the opportunity to understand how driving simulation technology can be integrated into the highway design process without the need for significant changes to corresponding and associated business processes. As a result of the novelty of the process, although this paper focuses on findings and recommendations for the USH and Watertown Plank Road interchange, sections of the paper address how the methods and lessons learned from the research can be used to evaluate future projects at different stages of the design. Paper Structure and Outline Due to space limitations this paper presents an abridged version of a report prepared by the authors. Readers interested in more details about the project are encouraged to contact the corresponding author with questions. Besides the introduction, the paper is divided into five main sections: Simulation Model and Scope, Early Benefits, Experimental Methodology, Results, Transferability of Findings, and Conclusions. Worthy of noticing about the aforementioned section is that in the Simulation Model and Scope section readers are provided with an overview of the driving

4 Santiago-Chaparro, Reichl, Bill, and Noyce simulator scenario used and areas of the model that were of interest to the team and were therefore the focus of the experiments. Also, the Early Benefits section provides a summary of how the design team used the scenario created by the research team as part of their design process even before actual subject experimentation took place. The Results section is structured in a findings and recommended actions structure meaning that findings for a particular area of the design are discussed and then a recommended solution is provided. A final section with content that is worth mentioning ahead is the Transferability of Findings section. As previously mentioned, while the authors present the findings of a driving simulator study on a particular project, the aforementioned section provides guidance on how the procedures used can be used to evaluate future highway projects. SIMULATION MODEL AND SCOPE Figure displays a schematic top view of the interchange design studied in the driving simulator. The numbers on the figure identify areas of interest in the scenario that were identified as worthy of evaluation during the experiment. These areas were the result of discussions between the design team and the research team during the driving simulator scenario development process and were the result of several meetings and conference calls that took place with WisDOT officials and the design team to ensure the model in the simulator was representative of the proposed interchange design. During several of these meetings, WisDOT personnel and the design team were provided an opportunity to operate the driving simulator and drive through the simulated interchange scenario. These meetings turned out to be a crucial part of the evaluation process and a summary of the results of the meetings is provided in the Early Benefits section of this paper. Signage for Milwaukee Regional Medical Center As drivers travel southbound on USH (USH), the best way to access the Milwaukee Regional Medical Center and the hospital is by exiting towards eastbound Watertown Plank Road. As a result, the design drawings call for several signs that assist drivers trying to navigate to the Medical Center. Figure displays the proper route to the Medical Center, as well as the proposed signage near the southbound USH exit ramp. Analyses were conducted to determine the impact that the signs had on directing drivers towards the Medical Center. Signage on Watertown Plank Road for US Highway Entrances Access to USH from Watertown Plank Road was another area identified as warranting analysis. Entrance ramps for USH are accessed from the right lane of Watertown Plank Road. To support that message to road users the design relies on signage placed in the road median (as shown in Figure ) alerting drivers that all freeway entrances are accessed through the rightmost. Analyses were conducted to determine the impact that signs in the median had on directing drivers to USH. Pedestrian Warning Signs Located on USH Exits Drivers exiting the USH highway in the southbound direction towards Watertown Plank Road (eastbound) faced a cloverleaf-style curve containing a pedestrian and bicycle crossing at the end as shown in Figure. Analyses were conducted to determine how proposed signage and supplemental technology provide support in conveying the message to drivers that there is the possibility of pedestrians ahead once they finish navigating the exit.

5 Santiago-Chaparro, Reichl, Bill, and Noyce End of Scenario* Start of Scenario* N Ramp WA Watertown Plank Road Ramp WJ US Hwy Exit Ramp US Hwy Entrance Ramp Direction of Travel Figure. Overhead View of Interchange Figure. Route and Signage Plan for Milwaukee Regional Medical Center

6 Santiago-Chaparro, Reichl, Bill, and Noyce Figure. Signage for US Highway Entrance from Watertown Plank Road RRFB US Hwy Exit Ramp US Hwy Entrance Ramp Location of Crosswalk N Sign & Station WG RRFB RRFB = Rectangular Rapid Flashing Beacon Sign & Station. WG Watertown Plank Road RRFB Sign & Station WG RRFB Sign 0 & Station WG Figure. Pedestrian Signage on US Highway Exit Ramps

7 Santiago-Chaparro, Reichl, Bill, and Noyce EARLY BENEFITS The ability of having a realistic visualization of a proposed highway design in a driving simulator environment proved to be a valuable review tool for designers. As a result of the opportunities that designers had to visualize the design prior to subject experimentation during the scenario development process changes were made to the proposed interchange design and signage plan. A summary of the changes made is shown below. In order to provide additional time for drivers to process US entrance signs from Watertown Plank Rd the design team decided to move the location of a sign truss on the US entrance ramp approximately 0 feet upstream of the previous location. The aforementioned sign was also rotated degrees to align better with the path followed by the driver as they enter the ramp. Add a sign truss on USH entrance ramp from eastbound Watertown Plank Road because the design team realized that the sign was not included in the plans. Reduce the lengths of a barrier wall on the exit ramp from USH towards Watertown Plank Rd in order to increase the visibility of pedestrians walking on the sidewalk along Watertown Plank Road. Move ramp metering signs upstream of their previous position to reduce the number of yellow signs clustered near the start of the entrance ramps in order to reduce driver confusion. These aforementioned changes are typically the type of changes that can result in expensive change orders if identified during the construction phase and that can result in temporary closures id identified after the highway becomes operational. These changes identified by the design team were included in the driving simulator model that was used for subject experimentation, approved by the designers, and eventually made their way into the final design plans. METHODOLOGY A total of subjects were recruited to participate in the study. The study was approved by the institutional review board (IRB) of the University of Wisconsin-Madison. The approximate average driving experience of the subjects was years. Gender was close to evenly distributed with female subjects and male subjects. The subjects were involved in an average of 0.0 crashes in the past three years and received an average of 0. citations for moving violations. Upon arrival to the driving simulation laboratory subjects were provided with a brief explanation of the experiment and were asked to read a consent form. After reading the consent form, subjects were asked if they had questions. If the subjects had any questions these were addressed before moving ahead with the experiment. If a subject agreed to proceed with the experiment after reading the consent form and receiving an answer to all the questions the subjects were asked to sign the consent form and complete a demographics questionnaire that included questions about personal driving behavior. After completion of the questionnaire, the following steps were followed: Install the eye tracking equipment on the subject. Initial adjustment and calibration process of the eye tracking equipment. The subject was asked to enter the vehicle and adjust the seat to a comfortable position.

8 Santiago-Chaparro, Reichl, Bill, and Noyce Eye tracking equipment used for the experiment was the head mounted type in the form of a pair of eyeglasses. The eye tracking equipment was manufactured by ASL and sold under the name of the Mobile Eye XG. The lenses on the eye tracking glasses were unshaded and clear. No vision was obstructed by using the glasses. Upon initial installation of the eye tracking equipment the subjects were accompanied to the vehicle and asked to adjust the seat to a desired position. Once the subject adjusted the vehicle seat to a comfortable position, a training scenario was started. The goal of the training scenario was to make the subjects familiar with operating the driving simulator in a moving environment. Specifically, users became familiar with the steering wheel, as well as braking and accelerating maneuvers. As part of the familiarization process, the subjects were asked to perform the following tasks until the experimenter deemed that the subject was confident with the simulator: Join a freeway traffic stream moving at approximately MPH from a completely stopped position on the right shoulder. Change lane positions and maneuver between vehicles. Pull to the side of the road and join the traffic stream again. Steer through normal conditions and through curves. Once the researcher deemed that the subject was comfortable with the simulator and if no indication of potential simulator sickness were present the training scenarios was stopped and a final eye tracking equipment calibration was performed. During the final calibration process, the subjects were asked to look at certain points on the screen to create a reference frame that allowed for the monitoring eye movement using the eye tracking equipment. After the final calibration was completed, the experimental scenario (Watertown Plank Road and USH interchange) was loaded and the subjects were asked to navigate through the interchange and complete a series of navigational tasks. These tasks involved finding the location of the medical center, joining the USH, and exiting the USH. The experimental process (each subject run through the scenario) took approximately minutes. Upon completion of the experimental process the eye tracking equipment was removed from the subject and the subject exited the vehicle. After exiting the vehicle the subject were escorted to a debriefing station and asked to complete an exit interview. The purpose of the exit interview was to receive feedback from the subjects about situations in which they were confused as well as to get feedback from them about the experimental process. The debriefing process played an important role in the experiment since it provided designers from potential end users of design. RESULTS AND RECOMMENDATIONS Data collected from the study were analyzed using quantitative and qualitative methods. Based on the results, and by considering the schedule of designers and project managers, the following recommendations shown as subsections ahead were made. The authors understand that the experiment presented is a case study of particular design and as a result the next section of this paper discusses how the techniques used in this experiment can be used in other projects. Furthermore, results are presented in non-technical terms (to the extent possible) in order to appeal to a broader audience interested in how driving simulator can be used to evaluate proposed highway designs instead of for what can be considered fundamental research.

9 Santiago-Chaparro, Reichl, Bill, and Noyce Signage for Milwaukee Regional Medical Center from Southbound US Highway In order to study the effect that the medical center sign located on the USH southbound direction has on the navigation task, a preliminary scenario in which the sign was removed was used. Only three subjects were exposed to this scenario as a method of establishing a base line given the time limitations. The instructions given to the subjects were to start driving and to use the exit towards the medical center. All the subjects missed the exit. The remaining subjects were exposed to a scenario containing the aforementioned missing sign and were given the same set of instructions. The majority of the subjects correctly identified the sign and positioned themselves in the right lane and started looking for the medical center exit. However, several subjects were confused when faced with the sign indicating the 0A-B exit and returned to the center lane in order to continue looking for the Medical Center. Some additional instances in which the sign was missed by a subject resulted from a heavy vehicle blocking the view of the sign. During the exit interview subjects who were confused about the location of the medical center exit were asked about their experience. Some pointed out that they correctly identified the sign but were confused because no additional indications about the location of the Medical Center were provided and they were looking for an additional sign. Recommendation: Based on the observed eye tracking behavior and exit interviews, the research team recommended adding an additional sign for the Medical Center. The additional sign will assist drivers in the case that they miss the first sign or are confused about their previous lane selection. Eye tracking data suggests that the majority of the subjects fixated on the overhead sign located by exist leading to the Medical Center; therefore, the overhead sign provides a good placement opportunity for the additional sign. While the MUTCD and State signage policy does not allow for two signs for supplemental signs on the freeway either through an exemption from the policy or through another approach drivers should be advised more clearly about the location of the medical center. Pedestrian Signage on US Highway Exit Ramps to Watertown Plank Road Available eye tracking data (sample shown in Figure ) from subjects that correctly navigated the ramp with the pedestrian warning signs shown in Figure suggest that there are no issues with identifying the pedestrian ahead warning sign located immediately at the beginning of the exit ramp. In fact, the data suggests that on average, subjects fixated on the sign prior to reaching the location of the sign. On the other hand, the data suggests that after the initial fixation on the previously mentioned warning sign the subjects do not fixate on the pedestrian crossing area located at the end of the exit ramp until approximately to seconds before arriving at the crossing area. Recommendation: Based on the eye tracking behavior and the small time difference between fixation on the crossing and arrival at the crossing, the research team recommends that additional awareness be created about the possibility of a pedestrian presence at the end of the exit ramp shown in Figure. The additional awareness can be created through the use of Rectangular Rapid Flashing Beacons (RRFP). These beacons were already under discussion before the simulation experiments. Therefore, the findings confirmed the benefits that the beacons could provide.

10 Santiago-Chaparro, Reichl, Bill, and Noyce 0 0 Signage on Watertown Plank Road for US Highway Entrances During the experiment, subjects were asked to go toward USH N/S while driving on Watertown Plank Road. Instructions provided to the subjects were to go towards USH N/S while driving on Watertown Plank Road before any of the USH N/S signs were visible. A portion of the subjects were instructed to stay in the left lane and others were not provided any instructions. Subjects who were instructed to stay in the left lane correctly identified that they needed to move to the right lane after noticing the signs and proceeded to exit correctly. The subjects who were not given specific instructions also identified the signs and proceeded to exit correctly. However, during the exit interview, several subjects indicated they were confused by signs that were placed on the left side of the road instructing and instructing them to be on the right lane. Recommendation: Based on the exit interview process and by observing eye tracking data, the research team recommends that signs providing information about the proper lane for accessing USH N/S be supplemented with an arrow pointing diagonally to the right. 0 Figure. Sample Eye Tracking Data TRANSFERABILITY OF FINDINGS The focus of the project was to explore how a driving simulator can be used to evaluate the proposed design of the Watertown Plank Road interchange. However, a common question that the research team received during the project was how the procedures used can be applied to other projects and the effort required given the desired results. Figure describes the benefits observed after conducting the evaluation of the aforementioned highway project and the information that could be obtained for other projects at different stages of design, i.e., 0, 0, and 0 percent plans.

11 Santiago-Chaparro, Reichl, Bill, and Noyce As the design becomes more complete, the level of testing that can be performed becomes more sophisticated. However, at the same time the window of opportunity for making significant becomes smaller. Therefore, there are different benefits that can be obtained by evaluating a proposed highway design during different stages of the design process. During the early stages of the design (0% plans), the simulator environment can be an invaluable tool for visualizing geometry and testing different alternatives. In fact, when members of the design team for the USH and Watertown Plank Road interchange used the simulator to visualize the design before it was ready for subject testing, the potential for understanding the design and proposing changes was immediately realized. During the visualization process, designers were able to identify changes that should be made to barriers and the position of signs, among others. With an additional level of effort, during more advanced stages of the design process (0% plans) the simulator can be used for more complex evaluations. Not only is studying the geometry of the design possible but different construction staging alternatives can be studied. The evaluation of construction staging alternates could help with the identification of road segments where drivers are potentially overwhelmed with the roadway configuration. Once these potentially overwhelming segments are identified, designers can use the information to modify the construction staging to promote, to the extent possible, an optimal driver workload. Furthermore, at this stage of the process designers can use the simulator to start looking at roadside objects positioning. A review of roadside objects positioning can ensure that sight is not obstructed as a result of the vehicle cabin characteristics and the proposed design. The same review can be used to ensure that the navigation task is not hindered by roadside objects. Finally, at the most advanced stage of the design process (0% plans), which requires the highest level of effort, is where the opportunities for conducting significant subject-centered testing appears. While at this stage of the process there is an opportunity to understand the comprehension and behavior of subjects as they navigate the design, changes to the design resulting from findings can be limited due to the increased cost of change. Therefore, changes at this stage of the design might be limited, except in the case of critical findings, to signing and pavement marking modifications, and the addition of supplemental safety equipment such as speed feedback signs, RRFP, among others. Nevertheless, it should be noted that significant changes remain possible and are a better alternative compared to design changes during construction. CONCLUSIONS This paper presented the results of a simulator-based evaluation of different design aspects of the proposed US Highway and Watertown Plank Road interchange in the vicinities of Milwaukee, Wisconsin. Through the use of state of the art driving simulation the research team showed the feasibility of merging the latest design practices with the latest advances in technology to bring road design and safety evaluations to the st century. Using a full scale driving simulation needed changes to the proposed highway design were made before final plans were complete. More importantly, changes were identified before the start of construction. Examples of need changes identified by the research team, working in conjunction with the project design team, included the identification of structures requiring relocations, missing signs, and messages in need of reinforcement.

12 Santiago-Chaparro, Reichl, Bill, and Noyce Analysis Options and Effort Requirements 0% Plans 0% Plans 0% Plans Low Effort Medium Effort High Effort Design/Signing/Marking Changes Testing from a Human Factors Perspective Geometry Visualization Construction Planning Design Fine Tuning Sign Comprehension and Final Signing Plan Alternative Testing Roadside Objects Positioning Project Benefits Final Marking Tweaks and Design Identified the Need for Overhead Sign Relocation and Rotation Identified the Need for an Additional Medical Center Sign Identified the Need for Additional Pedestrian Presence Awareness Identified the Need to Supplement Proposed Signs With Additional Information 0 Figure. Uses of Procedures Beyond the Scope of this Project Financial and safety benefits of using full scale driving simulation to evaluate future designs are clear since a safer road can be provided and the number of change orders needed during construction is reduced. However, exploration of the potential of using full scale driving simulator for proposed highway design evaluations have only begun. Potential benefits from the methodologies outlined in this paper are widespread. By studying the driver behavior on a proposed road project designers can produce a safer road for drivers; a road that provides optimal support for the control, guidance, and navigation tasks. In addition, there are vast financial benefits

13 Santiago-Chaparro, Reichl, Bill, and Noyce 0 by identifying problems before construction begins thus avoiding the cost of change orders or the reconstruction of project segments. The benefits of such a study far outweigh the costs in the long term and governing bodies should consider the methodology benefits when evaluating projects of a significant financial impact. Additionally, the benefits of the driving simulator extend beyond the evaluation of a particular project whether financially significant or not. The driving simulator can be used to help designers and transportation agencies test design alternatives/solutions that are not standard or fall outside of the recommended design guidelines parameters. And while non-standard designs/solutions are not a designers first-choice these deviation from guidelines are increasingly needed in this day and age as a result of increasing the complex environmental regulations and land availability. Therefore, driving simulators are not only a tool to improve the safety of a project but also a tool that can help in the development and testing of new design practices.