Freightliner IVI FOT
|Freightliner IVI FOT|
|Type: Field operational test|
|Tested system/service: Autonomous Systems|
|Countries: USA||? test users|
|6 partners||100 vehicles|
|Active from 2000 to 2001|
|Data catalogue||Tools catalogue|
|Data sets used in this FOT:
||The following tools|
were used in this FOT:
Freightliner tested two related but distinct functions of a new Roll Advisor and Control (RA&C) system. The RA&C is designed to assist commercial vehicle drivers, especially drivers of tanker trucks, in avoiding rollover crashes. The benefits of RA&C for helping to avoid single-vehicle roadway departure (SVRD) crashes were also evaluated. During the time of the FOT, Freightliner began offering the RA&C for sale.
The first component is the Roll Stability Advisor (RSA), which is intended as an educational tool for drivers. The RSA will not prevent any particular crash through direct intervention. Instead, it merely advises a driver, after a maneuver is finished, that the lateral forces on the vehicle were higher than might have been desirable. The RSA’s advisory notices are provided to the driver as briefly worded messages appearing on the instrument panel display. The second component is the Roll Stability Control (RSC). This system takes partial, momentary control of the vehicle if it deems that a serious rollover threat is developing. The system’s authority in the FOT was limited to reducing the throttle or applying engine braking. Only a minimal amount of deceleration is applied in this manner, but the hope is that a bad situation can be prevented from becoming worse.
As the independent evaluator, Battelle and a team of subcontractors analyzed the FOT data and conducted independent studies to estimate the safety benefits of the RA&C, assess driver acceptance of the new technology, and study the prospects for widespread deployment. Battelle participated in driver interviews and conducted independent track tests of the RA&C to support a system performance assessment and to validate the vehicle dynamic simulation models used in the safety benefits estimate.
Source: final report
Details of Field Operational Test
Start date and duration of FOT execution
September 2000 – December 2001
The Experimental Design Called for Two Periods: during the Baseline Period, data was recorded on Normal Driving; during the System On Period, data was recorded to look for any effects of the Roll Advisor & Control system (RA&C) Data analysis and results were completed in November 2002.
Indiana, Michigan, Illinois, and Wisconsin
In 1999, USDOT entered into cooperative agreements with four partnerships to conduct Generation 0 Field Operational Tests (FOTs) of advanced intelligent vehicle safety systems (IVSS). These systems are expected to begin production preparations before the end of fiscal year 2003.
Although the scope of the IVI Generation 0 FOT program includes light passenger vehicles and transit vehicles, USDOT selected one FOT involving specialty vehicles and three FOTs involving commercial trucks:
- Minnesota DOT tested technologies designed to provide operators of snowplows, ambulances, and patrol cars a means to maintain desired lane position and avoid collisions with obstacles during periods of low visibility. Key among these technologies are vision enhancement, lateral guidance, and collision warning systems.
- Volvo Trucks North America, Inc., in partnership with U.S. Xpress, tested a forward collision warning system, a blind spot warning system (not under evaluation), an adaptive cruise control, and an advanced electronic braking system for commercial vehicles
- Mack Trucks, Inc., in partnership with McKenzie Tank Lines, tested a trucker safety advisory system and a lane departure warning system for commercial vehicles
- Freightliner Corporation, in partnership with Praxair, tested a roll stability advisor and a roll stability control to assist commercial vehicle drivers in avoiding rollover crashes.
The primary evaluation goal of the FOT was to determine the potential safety benefits of IVSS. Specifically, how many crashes, injuries, and fatalities could be avoided if all such vehicles were equipped with these technologies? It is also important to understand how these technologies affect driver performance. For example, do drivers drive more safely? And how do these technologies affect driver stress level and workload? The secondary goals of these evaluations include the estimation of other benefits (mobility, efficiency, productivity, and environmental quality), evaluation of system performance, and assessments of other factors that affect development and deployment of these technologies. These factors include user acceptance, product maturity, manufacturability, and institutional and legal issues.
Goal 1: Achieve an in-depth understanding of the benefits of intelligent vehicle safety systems (IVSS). Because the benefits of IVSS fall into five different categories (safety, mobility, efficiency, productivity, and environmental quality), this goal area is divided in five separate goals, each corresponding to a different benefit category.
Goal 2: Assess user acceptance and human factors. Evaluation of driver performance (in particular, answers to the question: “Do drivers drive more safely with IVSS?”) are considered under Goal 1A – Safety Benefits.
Goal 3: Assess IVSS performance and capability potential.
Goal 4: Assess product maturity for deployment.
Goal 5: Address institutional and legal issues that might affect deployment.
For more details see page 3-1 of the final report: http://www.itsdocs.fhwa.dot.gov/JPODOCS/REPTS_TE/13871/13871.pdf
The RA&C was still in a developmental stage when it was deployed in the Field Operational Test. The drivers participating in the test were skilled and experienced. That the advisory function had any effect at all in improving drivers’ practices bodes well for its larger deployment. Even in this study of 12 months, 15 drivers, and only six tractors, a statistically significant reduction in risky driving behavior was observed. All parties, including Battelle (the independent evaluator) and the drivers themselves, believe its benefit will be greater when it is deployed more widely, particularly with less experienced drivers.
The FOT raised a number of questions, and suggestions for addressing them. Meritor-Wabco has continued to work on the RA&C since it was deployed in June 2001, so some of the points noted in the final report, particularly those concerning improvements to the device, may already have been addressed.
Just as the technology of the RA&C has developed between 2000 and 2002, Freightliner and Meritor-Wabco’s plans for the device have evolved since the application for a cooperative agreement was submitted. The most prominent change has been the shift in focus from the purely advisory function of the RSA to the intervention function of the RSC. The larger portion of the safety analysis concerns the RSA’s effect on driving behavior. The safety benefits estimate and the benefit-cost analysis are both based entirely on changes in driver behavior.
The Field Operational Test was efficient and effective. Communication between Battelle, the independent evaluator, and its primary contact in the partnership, UMTRI, was excellent. Staff at the DOT was willing to provide guidance or assistance when necessary.
An important factor contributing to the success of this FOT was the existence of a sound experimental design and data acquisition plan at the early stages of the project. Measures of safety were considered at the outset, and the data were rich enough to allow other questions to be answered as analysis progressed. The Partnership’s plan to collect 100 percent of the driving data, rather than triggering data collection based on pre-specified driving events, made it possible for the independent evaluator to adjust threshold values and perform special queries as the analysis progressed. This approach ensured that the data analysis and findings would not be adversely affected by the data collection process. The ability to raise and answer new questions proved invaluable. There were many parts of this report where an analysis could be conducted only by studying all trips through a given road segment, not just those trips where a pre-defined trigger condition was met. Continuously recording helped to improve the quality of the data itself, too. The drift in the drive axle accelerometer needed to be corrected more often than UMTRI anticipated. Because data were recorded on long, straight stretches, the signal could be zeroed according to the known cross slope of several straight highway segments.
The use of a single terminal kept logistics to a manageable level. There were two occasions when a Battelle staff member interviewed the drivers. In both instances, nearly all of them were seen in a single trip of only a couple days. A benefit of the limited delivery area was that there were many road segments through which each driver passed many times. There were, however, disadvantages to the selection of a single terminal. Only six tractors, driven for six months each in the control and treatment phases, produced a fairly small number of serious events to analyze. By the nature of statistics, confidently estimating small changes, as were expected and observed in this FOT, requires large numbers of near-rollover incidents, which were simply not available. In this sense, the compact size of the FOT fleet was largely responsible for the lack of statistical significance in the final safety benefits estimate. Also, while the convenience of the terminal’s proximity to both UMTRI and Battelle had its benefits, the RA&C was tested exclusively in the Great Lakes region of the country, which is fairly free of hills and curves. As with many experiments, there is an inherent trade-off: a larger experiment provides richer data but a smaller one is easier to conduct.
Data collection was largely automatic. UMTRI reviewed the data to make sure all was working well. On ordinary days it took no help from Praxair personnel. In the cases where valid data were missing or the data collected were invalid, the partnership could readily trace the path through the server at the terminal to diagnose and fix the problem.
There were two ways that the experimental design limited the strength of the conclusions. The FOT was conducted essentially within a one-year period, with one phase conducted as the weather was warming and the other phase conducted as the days shortened. Separating the seasonal effects from the RA&C’s effects could not be done with absolute certainty. Secondly, the demographics of the drivers was uniform. All were experienced, and all were within a twenty-year age span. A better test of the educational benefits of the RSA would have been to study two similar groups of novice drivers, one with the system and one without. Battelle understands, of course, that any experiment must be planned within constraints, and the partnership worked around these constraints as well as possible.
Personal interviews with the drivers allowed a much greater understanding than would have been possible had all the human factors been collected with pencil surveys. Even during the final interview, which was quite structured, the extra comments of the drivers revealed insights on how they interacted with the system, what they thought of its behavior, and even how they drive their trucks.
The conclusions of the final report were strengthened by Battelle’s ability to test hypotheses and run experiments on the test track. This was possible only through the cooperation of all four members of the partnership. Praxair donated a trailer to the project and provided valuable information on its operations. Meritor-Wabco allowed Battelle to use its flatbed trailer and ballast for one of the tests, and it provided two RA&C units, one configured with RSC and one without. UMTRI allowed Battelle to use a DAS and several instruments for the experiments, so the test track data would be readily comparable to the FOT data. Freightliner leased a tractor to Battelle for the tests. Much of the delay in performing the evaluation was due to difficulties obtaining the tractor for the test track experiments. All parties were understandably reluctant to reveal sensitive corporate information, and, of course, not everything that Battelle requested was granted. The success of the project was possible only with the cooperation and open communication and the sense that all were working toward a common goal.
Summary, type of funding and budget
US Department of Transportation (USDOT)
Cooperation partners and contact persons
- US Department of Transportation (USDOT) – sponsor
- Freightliner – major partner
- Praxair – fleet operator
- University of Michigan Transportation Research Institute (UMTRI) – underlying technology, FOT logistics
- Meritor-Wabco – component supplier
- Battelle - independent evaluator of the FOT on behalf of the USDOT.
Main Contact person
Jerry Pittenger Program Manager Battelle 505 King Ave., Columbus, Ohio 43201-2693; Telephone: +1-614.424.5189 pittengerj[at]battelle.org
Applications and equipment
Roll Advisor & Control system (RA&C) measures the vehicle’s lateral acceleration, estimates its mass, makes some assumptions about the trailer, and then decides whether to advise the driver or slow the truck.
The Praxair fleet that participated in the FOT carries compressed gas in bulk liquid tank trailers. Nationwide, Praxair operates approximately 650 power units that average nearly 100,000 vehicle miles traveled (VMT) per year. The LaPorte depot, which participated in the FOT, operates 13 power units. The drivers for Praxair are considered to be among the safest and most skilled in the industry. Examination of carrier information from the Safety and Fitness Electronic Record (SAFER) system confirms the impressive safety record of Praxair and its drivers. Over the past 24 months, 850 roadside inspections were performed on Praxair vehicles with only 61 vehicles (7.2 percent) placed out-of-service (OOS). The national average OOS rate for vehicle inspections is 23 percent. During this same period, only four of the 1,171 Praxair drivers inspected (0.3 percent) received OOS orders – compared to the national averages of six percent for HAZMAT drivers and eight percent for all drivers. Praxair’s SafeStat Safety Evaluation Area (SAE) scores for vehicle and driver inspections are 13.66 and 5.37, respectively. The SAE score represents a percentile of the distribution of safety measures among motor carriers. For example, Praxair’s driver SAE score of 5.37 means that their driver OOS rate is lower than almost 95% of the carriers in the country.
Praxair’s accident SAE score is 9.16, which means that Praxair’s recordable accident rate is among the lowest 10% within the motor carrier industry. According to Praxiar’s own records, their fleet of 650 trucks average 5.8 rollovers per year.
Equipment carried by test users
UMTRI installed an automated data acquisition system (DAS) in each of the six FOT tractors. The DAS acquired data from several sources—vehicle condition from the vehicle data bus, location and time from the Global Positioning System (GPS), and several special transducers installed by UMTRI in the tractor.
Pre-simulation / Piloting of the FOT
Method for the baseline
Techniques for measurement and data collection
Historical population crash data came from the National Automotive Sampling System (NASS) General Estimates System (GES), and the corresponding fatality rates were derived from the Fatality Analysis Reporting System (FARS). Annual rates of crashes, injuries, and fatalities were based on averages for the years 1995 through 2000. The host fleet, Praxair, also provided crucial information on its own rollover experience. Had any crashes or incidents occurred during the FOT, they would have been investigated, but, as expected, there were none.
In addition, subjective data was collected through driver interviews and surveys as presented in the table below (source - final report[].
Recruitment goals and methods
Source: final report
Methods for the liaison with the drivers during the FOT execution
Methods for data analysis, evaluation, synthesis and conclusions
Five main sources of data and information were used to conduct the evaluation:
Historical and FOT Crash/Incident Data. This source includes any available databases on truck crashes and relevant incidents. Primary sources were public databases—such as Trucks Involved in Fatal accidents (TIFA) and the General Estimates System (GES)—and the fleet’s crash history. In the estimation of safety benefits, this represented the crash incidence and distribution “without” the safety system.
Onboard Driving Data. The heart of the FOT, this source includes all data collected on the vehicles during the FOT. It was studied extensively to determine how often and under what circumstances possible precrash conflicts occur. Critical conflicts identified in GES as being relevant to rollovers were sought, as well as any instance where the lateral acceleration approaches the rollover threshold.
Surveys and Interviews. Opinions were solicited from personnel in the FOT (including drivers, mechanics, and corporate staff) and used to determine whether the messages are clear and to gauge the level of user acceptance, product maturity, and institutional and legal issues.
Fleet Operations Records. The operator’s maintenance and operation records that are relevant to the FOT were examined to help estimate the costs or savings associated with the IVSS.
Special Tests and Supplemental Data. This category includes all sources of data outside the FOT itself. The most significant of the special tests were test track maneuvers with an instrumented tractor and a trailer equipped with outriggers. Other sources of data were interviews with representatives of various stakeholder groups, to ask what institutional or legal issues might affect deployment.