When is a research product considered operational?

Variants of this question come at different stages of the R2O cycle following the initial transition of a research product. On its face, the term “operational” is vague within the context of R2O. If you open a dictionary, you will probably find that “operational” confers readiness to use. Because the R2O process is multi-faceted, that is, there are scientific and technical components, the threshold at which readiness is achieved will depend on the role and perspective of the stakeholder within the cycle. You may consider a new development to be operational once it is available to a subset of practitioners and those practitioners have been trained on how to use it. Others may require availability throughout the organization.

A former Meteorologist-in-Charge (MIC) at the Milwaukee field office of the National Weather Service (NWS), Ken Rizzo, believed there were two basic ingredients to a successful transition to operations: consistency and reliability. Consistency refers to the requirement that there are no longer any major developmental changes and ensures the maturity of the transitioned research product; reliability captures the need for high availability. A new product is unlikely to gain usability within operations if it is not there when practitioners need it, or its behavior deviates significantly over time so that it is difficult to get a sense of how to apply it.

A basic rule of thumb is that a product is not operational if it does not meet Ken’s ingredients. But even if it does, there may be certain organizational requirements, related to the scope of availability, and real-time monitoring of product creation, that establish attainment of operational status.

For example, the NWS has a central dissemination system that is monitored around the clock to deliver weather data and prediction model output to field office users. This system itself is under the scrutiny of technical managers and the classes of data that flows over this system are at the advisement of the science managers. All data is served to all users. For that reason, data disseminated in this manner is generally considered critical for weather forecasting – supporting the operational mission. To draw a parallel to the Information Technology Infrastructure Library (ITIL) framework, this mission-critical system meets both utility (“fit for purpose”) and warranty (“fit for use”) requirements to support the organization.

But in the “big data” era, we must recognize that everyone may not need all data at all times. That is, it is possible to narrow our requirements of fitness for purpose and use to be specific to product, class, or, in the weather community, phenomena, instead of enterprise level. Some research and development may lead to products with certain niche applications, or those that are supplementary to basic observational data. The challenges associated with “big data” will require new strategies for developing configurable production systems the enable cost-efficient and timely R2O in achieving operational status.

As data requirements evolve and options for network growth expand, the movement of the operational target is likely to change. It is important that stakeholders agree upon the definition of “operational” at the onset of a R2O project, and identify where it fits within the cycle. But wherever that may be, achieving operational status is hopefully not the end. It is also necessary to keep in mind that just because something is ready for use, does not mean it is actively in use. The success proposition at the backbone of R2O transitions requires active use as an indicator of value.