Hierarchical Database and Tools for Relating Technologies to Mission Requirements for the New Millennium Program

Can a centralized database of mission requirements and technology capabilities help the process of making decisions about technology funding and selection?

Future NASA missions will rely on breakthrough technologies to enhance capability and reduce cost. But new, unproven technology is risky to use when its failure may jeopardize an entire mission. The New Millennium Program (NMP) removes that "risk of first use" by conducting dedicated tests of new technologies in space, so first use doesn't happen when the technologies are critical to the success of a science mission.

Since funds for space-testing are, of course, limited, it is crucial to select technologies for this program that would provide the most value to NASA Enterprises. Hence, in making its selections, NMP looks for common technology threads that apply to several different missions.

We are developing a hierarchical data base and a tool for using it, to help NMP identify the technologies best suited for testing in space. Ultimately, we hope it will also be useful for decision-makers across a broad range of programs and missions. The database is based on XML (eXtensible Markup Language). The tool is called "XCALIBR" for XML Capability Analysis LIBRary.

Taxonomy

The heart of the XCALIBR system is a detailed taxonomy, including a data dictionary, which rigorously defines the structure and content of the database. The database uses the taxonomy as a filing system for organizing mission requirements and technologies, in much the same way that libraries use the Dewey decimal system to organize books. The taxonomy provides the means by which the database defines qualitative data, quantitative data and relationships.

The database currently has a comprehensive taxonomy for spacecraft bus subsystems, covering everything from high system-level metrics to circuits and fasteners. Also included is a set of remote sensing and in situ instruments and their associated metrics.

The entire taxonomy contains almost 700 elements, with each element representing a particular spacecraft subsystem, component, or part. Each element contains a unique set of metrics that characterize performance. Thus, the XCALIBR taxonomy has sufficient breadth to cover an entire space segment and sufficient depth to specify performance metrics at the component level and below.

Developing and using standard XML taxonomies will allow NMP and other decision-makers to trace the connections between mission requirements and new technologies on a scale not feasible with spreadsheets and relational databases alone. If accepted and ratified by the appropriate government, academic, and industry standardization bodies, the XML dictionary will for the first time provide a common language, allowing free and unambiguous exchange of information across all sectors of the aerospace community.

While the idea of matching missions and technologies via a spreadsheet has been around for some time, the difference here is that this database will, if broadly accepted, serve as a central repository of information, maintained and kept current on a permanent basis. Its hierarchical character will make it much easier to "file" information in places with logical relationships to each other, to draw useful relationships between technologies and the missions whose requirements they may satisfy, and to allow the taxonomy to grow and evolve as needed.

At present, we've developed three taxonomies:

1. Organizational: decomposes NASA into its various science themes and sub-themes.

2. Functional: organizes all of the functions or operations that a space mission encompasses (e.g., launch and early orbit operations, cruise, science operations, entry and descent operations).

3. Structural: organizes all of a spacecraft mission's software and hardware components, including the ground system, launch vehicle, spacecraft bus and payload.

Using XCALIBR and the XML Database

Following is a description of the alpha-test version of XCALIBR and the XML database. Please note that the scenarios described here assume that the database has been populated with the relevant information about both mission requirements and technology capabilities. We expect to add significant capabilities to the system by the time it is beta tested.

The user makes a selection from the hierarchical tree of mission requirements at left. Here, the selected node is "TPF Formation Flying."

XCALIBR presents the user with an extensive hierarchical tree of missions and their requirements. The user selects a node, which can be as high-level as a family of missions, or as low-level as a specific bolt.

Note that the mission requirements node has been populated with desired metrics (such as range for a radar system or threadcount for a bolt). The user clicks on either "Find Technologies" or "Analysis" to search the database for technologies that might satisfy the requirement.

If clicking "Find Technologies," the user also selects a query level. Level 1 is a match of type (e.g., if the selected node is a type of range finder, a level 1 query will produce all range finders included in the database). Level 2 is a subset of level 1 in which none of the technologies' listed metrics conflicts with the metrics for which the user is searching. Level 3 is a subset of level 2 in which the search produces only technologies that list all of the searched-for metrics, and have the values specified by the requirement.

The results page lists all the technologies in the database that purport to enable the selected node, at the selected match level.

For example, if the node is "formation flying," with a Level 2 query, the results page will list all the technologies in the database that meet all of the following criteria: (1) purport to enable formation flying, (2) list one or more of the metrics that were included in the user's search, and (3) all of those matching metrics have values that match those desired by the user. Note that at this query level, the technologies on the results page do not necessarily include all of the metrics the user listed.

If, instead of clicking "Find Technologies," the user clicks "Analysis," the search engine will automatically perform both a Level 1 and a Level 3 query, not only for the selected node, but for all "children" of that node. In other words, the search engine will look at the selected node and the entire portion of the tree that branches off from it. The results page will be an Excel spreadsheet, such as the one shown here:

The entire hierarchy of mission requirements derived from the selected node are listed on the left, while the applicable technologies appear across the top of the grid. Level 1 matches are indicated by a green cell. Where there is a Level 3 match, the number "1" appears in the green cell.

Clicking on an individual technology listing in either kind of results page yields a page of further information on that technology.

Testing

We performed alpha testing of the interface with representatives of the user community. In addition, members of our team worked through a sample problem, using actual data that had been used to conduct return-on-investment analysis for large space telescope technologies.

This study compared 7 technology areas on the basis of their predicted improvement over state of the art (SOA). Calculating that ratio yields a unitless value that enables comparison of dissimilar technologies.

The above table illustrates the data used in the sample problem. The study actually employed 100 metrics, so a complete table would include 100 rows like the ones shown here. The study also included telescopes with apertures of 35m and 100m in addition to the 10m version represented here.

In all, the taxonomy for this problem included 7 technology areas, 23 sub-areas, and 91 performance parameters, with 8 metrics per parameter. To accommodate this data, we needed to make a temporary change in our database. Such customized modification will cease to be necessary if, as our efforts seek to facilitate, a standardized system for entering data becomes widely accepted and employed.

We also modified our analysis tool to support the complex algorithm employed in the original telescope study, in which the analysis is a function of five variables (SOA performance, target performance, SOA technology readiness level, target technology readiness level, and cost). Values were computed in log base 2, in a manner that tended to equalize technologies with different levels of detail. We will retain this new capability in our analysis tool.

Our results, illustrated in the above graph, matched the results of the original study.

Lessons Learned

The alpha testing and sample problem produced the following lessons, which are being brought to bear as we develop this system:

1. The system needs to be flexible enough to support multiple views of the data, and allow developers to add new views easily.

2. Testers had difficulty fitting vague, high-level mission requirements and technologies into a specific predefined taxonomy. We need to improve our method of instructing users about the organizing principles of the taxonomy, and provide better online guidance. A map function can help. At the same time, using a taxonomy forces mission designers to define their terms clearly.

3. A graphical user interface is not sufficient for loading a large data set. More sophisticated data import/export tools are needed.

4. The system must be able to handle multiple analysis tools, since no one tool is likely to be suitable for all purposes.

5. XCALIBR needs a well-defined process by which custom taxonomies can be normalized and migrated into the official taxonomy.

At some point, we plan to explore other valuation methods. We expect that the database will evolve to store state-of-the-art and cost data for technologies, and the analysis tool will evolve to use these data types for more sophisticated valuation methods. Eventually, the database will serve as a common repository for several different types of analysis tools covering areas such as schedule and budget planning, as well as space mission trade studies.