SAE Recommended Practice J2944, Operational Definitions of Driving Performance Measures and Statistics

SAE Recommended Practice J2944, Operational Definitions of Driving Performance Measures and Statistics
From Paul Green, University of Michigan
Pagreen@umich.edu

SAE Recommended Practice J2944, Operational Definitions of Driving Performance Measures and Statistics was published in June of this year. Anyone reporting research on driving performance, should be using the terms and definitions in that practice. SAE Recommended Practice J2944, Operational Definitions of Driving Performance Measures and Statistics was published in June of this year after eight years of development. This document pertains to vehicles that either are driven by people or are self driving. J2944 is based on foundational documents such as the AASHTO Green Book, the Highway Capacity Manual, and the Manual of Uniform Traffic Control Devices, supplemented by more than 300 references.

This 170+ page recommended practice defines more that 80 terms related to driving such as lane departure, roadway departure, gap, headway, time to collision, and so forth, terms that are commonly used in CPS studies. This practice was written to overcome the misleading, inconsistent, and confusing use of terms related to driving. In the extensive literature review that serve to spur the development of this practice, there were sometimes more than 10 names for a common term (a lane departure was called a lane bust, lane exceedence, over the edge line, etc.) and many terms were defined 10-15% of the time, often inconsistently. For example, many crash avoidance studies incorrectly use the term headway. Headway is commonly the front bumper to front bumper distance (though there are actually two other alternative definitions). Gap, is intended measure, is the front bumper to rear bumper distance, 55 feet less than "headway" for a tractor-trailer.

To provide for consistent application, this practice recommends a single name for each term. However, recognizing the need to accommodate varying applications, many terms have multiple definitions. For example, for lane departure, there are 11 alternative definitions. Accordingly, in a publication, the following text could appear, "The dependent measures was the number of lane departures, option D, per SAE J2944." When measures and statistics are so specified, there is no ambiguity about what is reported, which is not the case now.

In addition to providing definitions, the document identifies where each term first appeared, if known, key studies in which the measures was used, guidance concerning its use, and in many cases, statistical distributions of each measure or statistic from on-road driving. For many, this information will serve as the basis of a literature review for each measure, and saving time. Graduate students will love it.

Researchers should expect to see that conferences and journals will require that authors cite the definitions in SAE J2944 or provide alternative definitions of equal specificity. Failure to do so could be grounds for rejecting a manuscript. This may seem harsh, but the current situation can lead to research that is not replicable, and not up to accepted professional standards.

Those interested in obtaining a copy should go to the SAE web site (SAE.org) and in the search box, type in J2944. That action will bring up a summary of this practice. Clicking on the practice title will bring up the page where one can purchase a copy ($72). We wish we could distribute free copies, but SAE is fiercely protective of its copyright. Think of purchasing a copy as being like buying an inexpensive but valuable book.

Those interested in additional information relating to this practice should contact Paul Green at the University of Michigan, the lead author. Suggestions for improvements, documents to cite, and other information are welcome. Contact Paul at Pagreen@umich.edu or call +1 734 763 3795.