nizations. Elements of that type of plan are outlined in Chapter 7.
• Establish a risk-based process for setting priorities among annual maintenance and repair activities in the field and at the headquarters level. Guidance for doing that is contained in Chapter 7.
• Establish standard methods for gathering and updating data to provide credible, empirical information for decision support, to measure outcomes of investments in maintenance and repair, and to track and improve the results.
Recommendation 4 (Finding 6). Federal facilities program managers should plan for multiple internal and external communications when presenting maintenance and repair requests to other decision-makers and staff. The information communicated should be accurate, acknowledge uncertainties, and be available in multiple forms to meet the needs of different audiences. The basis of prediction of outcomes of a given level of investment in maintenance and repair should be transparent and available to decision-makers.
Recommendation 5 (Finding 7). Federal agencies and other appropriate organizations should continue to collaborate to develop and refine governmentwide measures for outcomes of maintenance and repair investments and to develop more standardized practices, unambiguous procedures, definitions, and models. The committee believes that those activities would be most effective if under the auspices of the Office of Management and Budget.
Recommendation 6 (Findings 6 and 8). Federal agencies should avoid the collection of data that serve no immediate mission-related purpose. Agencies should use a “knowledge-based” approach to condition assessment. Outcome metrics and models should make maximum use of existing data. When new or unique data are required to support the development of an outcome measure or model, there should be a clearly defined benefit to offset the cost of collecting and maintaining them.
Recommendation 7 (Findings 8 and 9). Federal agencies should continue to participate in and take advantage of collaborative efforts to develop rapid and effective data-collection methods (such as the use of sensors and visual imaging devices), to develop data-exchange standards that allow inter operability of data and software systems, to develop the empirical information needed for robust prediction models, and to develop practices that will reduce the cost of data collection and eliminate human error and bias.