The University of Tennessee, Knoxville


Bicycles vs Railroad Crossings – A Case Study of Crashes

Source: Factors influencing single-bicycle crashes at skewed railroad grade crossings Ziwen Ling, Christopher Cherry, Nirbesh Dhakal.

More travelers are choosing bicycles as a healthy, economical, and environmentally-friendly mode of transportation. As bicycle travel increases, we must not fail to consider this mode of travel in ongoing transportation planning, operations, and safety activities. In some instances, bicycle-specific issues may not be immediately apparent. University of Tennessee Civil and Environmental Engineering Professor Chris Cherry, along with graduate students Ziwen Ling and Nirbesh Dhakal, studied one such bicycle safety issue that had previously gone undetected within view of the department's building on the UT campus.

Time Lapse of Bicycle Crash
Figure 1 - An example of a bicycle crash at the study location.

Video recordings of bicycles traversing a skewed highway-railroad grade crossing on Neyland Drive in Knoxville provided empirical data for this unique study of single-bicycle crashes. Observed crash rates were much higher than expected and almost all of them were unreported. The video was captured from the fourth floor of a building located approximately 450 feet from the crossing. The camera captured continuously at 720p resolution at 30 frames per second from August 2, 2014, to October 3, 2014. There are two cyclist grade crossings in this area; a parallel 10-ft. wide shared-use pathway (greenway) on the north side of the roadway (approximately 20° rail skew tangent to tangent), and a wide (approximately 11.5 ft.) shoulder on the south side of the roadway (approximately 10° rail skew). In both cases, cyclists can cross the rails at approximately 45° at low speeds within the shoulder or greenway. On the shoulder, many cyclists cross into the adjacent travel lanes to cross the tracks at larger angles. The speed limit on Neyland Drive is 45 mph.

During the study period, 13,247 cyclists traversed the crossings, including 9,521 on the Neyland Greenway (Eastbound and Westbound) and 2,091 on Neyland Drive shoulder (Eastbound), with the remaining 1,635 traversing at different parts of the roadway. During the study period, 53 crashes occurred at the crossing, including 21 (crash rate: 2.2 per 1000 traversings) on the Neyland Greenway (both directions) and 32 (crash rate: 15.3 per 1000 traversings) on Neyland Drive shoulder (eastbound).

Due to the restricted view and geometric characteristics, the study only focused on traversings and crashes that occur on the eastbound shoulder of Neyland Drive. That crossing is constructed from pre-cast concrete panels that includes a uniform 2.5-in. wide and 3–4-in. deep flangeway gap between the rail and the panel. An example of a bicycle crash mechanism at the railway crossing is shown in Figure 1.

Although approach angle is the most obvious and important factor to consider when designing rail crossings, other environmental or rider attributes influence crash outcomes. According to the study, group riders, women, and wet roadway conditions all contribute to higher crash rates. Facilities that share these characteristics should be designed with increased scrutiny.

Jughandle Crossing Design
Figure 2 – Jughandle design in the city of Knoxville

The study suggests countermeasures, like jughandle designs, that improve the traversing angle for cyclists. Improving traversing angle up to 90° is ideal but often infeasible due to space constraints. Any traversing angle more than 30° would be highly effective at reducing crashes and providing traversing angles greater than 60° would effectively eliminate them. The crossing studied here was infeasible to construct at a crossing angle that approached 90°. Responding to the problem, the City of Knoxville did construct a jughandle design (Figure 2) with a tangent angles of 57°, with a possible minimum angle of 37° (inside-to-outside of bike lane). This design encourages cyclists to approach this crossing with larger than 37° angle. It has effectively eliminated crashes except in cases where cyclists traverse the hash marks and cross at low angles.

A complete copy of the study is available at (

Please take a moment to watch a YouTube video documenting the crashes used in the paper. It is available at (


The University of Tennessee, Knoxville. Big Orange. Big Ideas.

Knoxville, Tennessee 37996 | 865-974-1000
The flagship campus of the University of Tennessee System