1.4.4.2 Performance measures

Each level of service should be supported by a set of performance measures (see figure 2), which will enable to measure both individual aspects of performance as well as the overall level of service.

Two type of measures are of interest:

• Measures of physical condition of the highway asset;
• Measures of the organizational performance.

It is relevant to distinguish between the condition and performance of an asset:

• Asset’s performance is directly related to its ability to provide the required level of service;
• Asset condition is related to its physical state, which may or may not affect its performance.

Usually, for each level of service, a number of performance measures are chosen, and the process to evaluate them generally involves the aggregation of raw data (objective and subjective, quantitative and qualitative), as shown in table 1.4.4.2.1.

Definition of Performance MeasureVarious performance measures are usually required to represent a single level of service, but they might be aggregated in a key performance indicator (KPI) to describe the level of service. It may be appropriate to use different weighting factors for each of the performance measures that contribute to a level of service or KPI (e.g. condition of footways may have a greater weighting than condition of white lines).

Typically, performance measures may include both engineering and non-engineering considerations (for example based on results from public opinion surveys). There are four type of performance measures:

• Input – Demonstrates the amount of resources needed to complete activities and includes staff, labor and material costs and other relevant financial information. This type of measure does not necessarily demonstrate the performance of the service or whether the overall strategy is met;
• Output – Typically measures completed activities, such as the area of surface dressing per year. This type of measure provides information on the activities completed against a program of works, but does not provide information on whether the level of activity met the overall strategy, whether it met any of the performance requirements, or if it delivered value for money;
• Outcome – Describes the results of activities provided and may include measures such as public satisfaction and the condition of various assets. These measures can be used to demonstrate the link between the asset management strategy and service delivery most effectively;
• Efficiency – Reflects the input cost per activity or for the total service. It could be used to support investment decisions.

The steps for identifying performance measures generally include:

• Identifying existing performance measures and how they are being used;
• Defining gaps to be addressed based on coverage of critical outcome areas for authorities/agencies and support for the asset management best practices;
• Establishing criteria for selecting new measures.

The logical starting point for the development of performance measures is to identify what aspect of performance is being measured.

Indicators suitable to be used as performance measures should be:

• Meaningful,
• Relevant to demonstrate the delivery of strategy,
• Based on reliable data already available,
• Simple to understand,
• Stable (can be compared over time),
• Useful for benchmarking (for comparison with other authorities/agencies),
• Already known or used (or something similar already being measured by others),
• Sustainable in terms of cost.

One of the basic considerations in developing performance measures is to maximize the use of existing data and minimize the need for collection of additional data.

The SMART approach may be used in selecting actual performance measures. This approach evaluates the following attributes for candidate measures:

• Specific - the measure is specific, clear and unambiguous, explaining clearly what is expected;
• Measurable - there is a simple method for measuring it;
• Attainable - the measure is realistic and there is a clear action plan of how it can be achieved;
• Relevant - it is important in achieving the asset management objective; and
• Time-bound it is measured in a timeframe.

Once performance measures are selected, it is advisable that:

• data are tested in order to verify if they are practical to measure and compute;
• validity of results is checked in order to verify that they genuinely reflect current performance (precision and accuracy of the performance measures is a critical issue when performance measures are established and target values set).

Typical sources of data and other information used to define performance measures may include: public opinion surveys (letters, e-mails, calls and blogs), business reporting, operational reporting, existing inventory data, performance data (including condition) and financial reporting.

In relation to KPIs, Working Group 2 (Management Indicators) of a past PIARC Technical Committee on Asset Management (cycle 2008-2011), put some effort into identifying the existing management indicators as well as the lack of indicators when necessary (PIARC 2012). Table 1.4.4.2.2 shows some examples of the work carried out by this committee.

Random Auditing: Audits can include customer satisfaction surveys, sample condition surveys and ad hoc inspections. Random audits are frequently used to independently audit the performance of Highway Authorities/Agencies who are working under quality management systems. Their purpose therefore is to provide an independent crosscheck on performance.As far as the frequency of measures is concerned, consideration should be given to the balance between the cost of collecting the monitoring data and information and the risks of not having the information available. This is particularly important when considering compliance with statutory obligations and demonstrating value for money. The most commonly used ways of carrying out performance measurement are:

• System Audits: Where asset management software systems have been implemented the system, if properly used, will provide a source of performance management data. For example, customer query systems will be able to provide response times to queries and also be able to identify areas of exception, i.e. particular types of complaint or query or particular geographical areas generating a high number of queries
• Monthly: Where established management systems are in place it is possible to obtain monthly performance statistics. In general, these may focus on activities and the time taken to complete them. Progress against an established program may also provide a useful management tool.
• Annual: Some indicators will be based upon information supplied by others annually. This limits the use of such indicators to annual reviews.