A harmonious platform for R2O
In a previous post, there was a discussion about how the United States federal government is instrumental in establishing and funding opportunities for R2O as a result of its position in the weather community. In this post, the conversation turns to whether it is possible to create a harmonious platform that enhances R2O and a partnership between the government, private sector, and academia. For this, we first examine the leading priorities of these three players and discern where there is alignment, and then discuss what needs to be done to better foster enterprise-wide collaboration. Because of the inherently diverse priorities among the three sectors, tensions can arise that can thwart opportunities that may be beneficial to the community as a whole.
The priority of the government is to save lives and property. This is done through better warnings and forecasts, and elected officials fund government agencies to ensure this public interest is satisfied. The challenge that arises is how the government should best spend its funding to meet their objective. What compounds this challenge is that there are often not “off the shelf” products that meet the sophisticated demands of a large operational entity, especially when it comes to observing systems of the regional and global scales. Without products to select from, the government is responsible for crafting requirements, showing purpose, and establishing use scenarios.
In doing this, there must be a significant degree of specificity. Often, in attempting to achieve this specificity, there is a lot of focus on hard operational benchmarks, particularly those related to information technology, such as limiting downtime, or meeting pre-established security standards, but not all benchmarks must be of this kind of guiding factor. It is easy to develop measurable benchmarks, particularly those that are simplistic. It requires substantial expertise and knowledge of a system of systems or other complex infrastructure to establish benchmarks related to data formats, system specifications, access privileges, and algorithm outputs, just to name a few. And beyond that, there must be some semblance of a vision as to how all of these details support the overall intent of the project, with traceability back to the government’s mission.
This can take a substantial amount of time to craft. The private sector and academic partners are generally more time conscious since they are predominantly task-focused, and more importantly, they have specialized expertise. Both are motivated to partner with the government as a customer. In contrast, though, the private sector is profit-minded as a first priority, and architects enterprise-level technical systems that meet the requirements of the customer, with interest in long-term operational support and maintenance of those systems. Academics tend to prioritize personal pride over profit in pursuing the right scientific solution to problems. They see the government as a joint stakeholder in the solution. Both can deliver technical solutions, but academics want to be able to test and update their code, maintaining ownership of the content and result, with less interest in a single centrally supported, operationally maintained deliverable. Scientists also largely share their research byproducts in open forums, such as journals and professional societies, whereas the private sector views the details of their solutions as proprietary.
Private sector entities feel competition from academia when large research centers with academic teams start to package their software and distribute it in an open-source arrangement. This is a challenging area. On one hand, the government paid for much of the algorithm development, and the algorithms are in the public domain. On the other hand, the private sector does not like the easy accessibility of these packages because they may appeal to prospective customers as a substitute for a more tailored solution. In some cases, however, packaged research algorithms do not intrude on market share. For example, international academic institutions and governments in developing countries generally do not have substantial budgets to cover operations-grade software and recurring support costs.
Given the relative strengths and priorities of the government, academia, and private sector, it is possible to envision a R2O platform for converting data to information that brings these partners together in harmony to support and advance the community, and potentially public, interest. The attributes of this platform are dependent on the partnership.
- The platform should provide access to developers of algorithms. While software standards are necessary, ownership of algorithms should remain in the hands of the scientists, not software developers. The private sector should resist trying to rewrite code unless critical to maintain it for an operational purpose. The vendor must work with scientists in improving code without imposing long transition timeframes. Operations and maintenance of code should be part of the contract with the vendor.
- The government must lead all governance activities, including establishing very clear software development standards. The private sector can provide recommendations in this area, but not in their realm of control.
- The platform should provide flexibility between operational, pre-operational, and experimental scientific products. There should be a test environment for scientists to run code, and an area the enables a broader distribution to users and partners, but does not interfere with the operationally essential products.
- The government should exercise great care in assuring that input and output formats, and access needs, are detailed for their requested solutions prior to awarding contracts.
Such a platform is clearly expansive in reach. Thus, it should not be lost that a substantial amount of vision is required to design a system that serves a purpose and meets a diverse set of needs. The government is central in accomplishing this, with the private sector providing technical acumen to the scientific interests from academics.
Latest posts by Jordan Gerth (see all)
- Preliminary, non-operational, but used operationally? Huh? - April 10, 2017
- Peer review and R2O - February 5, 2017
- Knowledge, skills, and mental representations - January 16, 2017