Inventory-Based Rating System

Inventory-Based Rating System:

A Stable and Implementable Assessment Method

for Unpaved Roads

Timothy Colling, PhD, PE; John Kiefer, PE; and Pete Torola, PE


Problem of Condition Assessment Systems for Unpaved Roads

While many condition assessment systems exist for unpaved roads, most of the systems evolved from paved roads. Consequently, assessment of road condition according to these systems relies heavily on surface distresses. The surface condition of paved roads significantly impacts road use by motorists; as such, an increase in surface distresses directly drives road-related improvements. Since the surface condition of paved roads has a slow rate of change (over the course of years), it serves as a good parameter for assessing and managing a paved road. As such, surface condition yields a network-level measure that can illustrate the impact of investments on a paved road network.

Unlike paved roads, the surface condition of unpaved roads can change rapidly (over weeks or even days) and is not always reflective of loss in road value. Unpaved road features like adequate ditches, culverts, minimum lane widths, shoulders, and sufficient structural gravel are more likely to influence road use. Furthermore, unpaved roads fluctuate greatly in design, construction, use, and upkeep.

Thus, assessment systems based on surface condition are problematic for unpaved roads.


Inventory-Based Rating System: The Premise

Three characteristic elements of unpaved roads are surface width, drainage adequacy, and structural adequacy. These three elements, in particular, have a substantial impact on road use and require a greater level of investment to create and rehabilitate. These elements do not change rapidly and, therefore, only require assessment when construction activities occur or when lack of maintenance leads to loss or degradation of the feature. As such, they can provide a network-level measure that illustrates the impact of investments on an unpaved road network. A baseline or good condition for each of these elements would be those characteristics that are considered acceptable by a majority of road users with guidance from design standards.

Therefore, the Inventory-Based Rating (IBR) System uses a baseline or good condition to assess unpaved roads, assigning an aggregate IBR score (using a nine-point system) based on individual ratings for Surface Width, Drainage Adequacy, and Structural Adequacy.



Applying the IBR System

Michigan’s unpaved roads vary greatly. Therefore, a pilot test project team defined three classifications of road networks: Low Volume Terminal Branch Networks (contain many "ends" of road systems where traffic volumes are low and few properties accessed), Agricultural Grid Networks (provide year-round regular access to farm fields and residents, supports higher traffic volumes and larger truck loads), and Suburban Residential Networks (provide local access to rural residential properties located near urban centers, supports primarily passenger vehicle traffic).

A sampling of five counties chosen for the pilot test reflect each of the different road network classifications. Data collection events in each of the counties included collecting IBR data and rating productivity measures, ascertaining gravel thickness by randomly performed measurements, and jointly collecting IBR data for unpaved roads and Pavement and Surface Evaluation Rating (PASER) data for paved roads. Data collection tools were Roadsoft and the Roadsoft Laptop Data Collector.

In general, the pilot test found that:

  • Low Volume Terminal Branch Networks exhibited narrow surface widths, minimal drainage adequacy, and minimal structural gravel leading to overall low IBR scores
  • Agricultural Grid Networks typically had good to fair surface widths, good to fair drainage adequacy, and good structural gravel leading to overall good or moderately good IBR scores
  • Suburban Residential Network had moderately poor IBR scores.

Rating productivity ranged from 6.3–8.8 miles per hour (10.1–14.2 km/hr) for Low Volume Terminal Branch Networks, which often had lakes and streams dividing the road network and/or contained many ends of the road network, to 28.3 miles per hour (45.5 km/hr) for good Agricultural Grid Networks, which featured roads in mile-long-section-line interconnected grid patterns; the average rating productivity was 12.3 miles per hour (19.8 km/hr)

Combined collection of IBR and PASER data had a higher rate of data collection (20.9 miles per hour, or 33.6 km/hr, in comparison to 8.8 miles per hour [14.2 km/hr] for IBR only or 14.8 miles per hour [23.8 km/hr] for PASER only) due to minimizing time travelled without rating

The IBR System had a high degree of accuracy (72.2 percent of "blind" ratings were exact matches with "ground truth" and 92.9 percent were within a tolerance of plus/minus one rating point; 79.6 percent of gravel thickness estimates based on institution knowledge matched the bin ranges established by actual measurements) indicating repeatability in the use of the system.

Currently, the IBR system is being taught in Michigan and used by state and local road-owning agencies.


For More Information

For more information on the Inventory-Based Rating System, see the article Inventory Based Rating System: A Stable and Implementable Method of Condition Assessment for Unpaved Roads by Tim Colling, John Kiefer, and Pete Torola. This article and this web page content were published as part of the Transportation Review Board's compendium of papers for the 2017 Annual Conference and as a TRB poster presentation, respectively.

Also, see the the article New Inventory-Based Rating System Pilot Tested in 2016 in the winter 2016 issue of Crossroads, the quarterly journal of the County Road Association of Michigan.

Follow we will begin offering training in the principles and application of the Inventory-Based Rating System in late summer 2017!

Questions? Contact us at or (906) 487-2102.