Automation has become a hot topic within the Department of Defense, but is it hot enough?
While the U.S. Marine Corps and the U.S. Army are investigating ways to include automation in battlefield scenarios, automation, as a rule, is still not generally regarded as highly as cybersecurity or even software development. Each of these disciplines typically has substantial budget line items and entire teams dedicated to implementation, a status that automation does not traditionally enjoy within the Pentagon.
And yet, automation should be considered just as mission-critical as both cybersecurity and software development, especially since automation drives both of these initiatives. Therefore, the Pentagon must raise the level of automation from something that agencies do “when and where they can” to a core discipline that takes place at the beginning of every project. As of yet, this has not happened to a great extent, for a few reasons:
First, legacy thinking still too often creeps into development processes, causing teams to implement automation only as necessary or in silos within different agencies. Second, there is still fear, uncertainty and doubt around the term automation, causing some people to fear being replaced even though true automation is about enablement, not replacement. Third, agencies that have tried automation in the past and have not succeeded — either due to a lack of tools or expertise — are hesitant to try it again.
These challenges are certainly not unique to the DoD. Nevertheless, they threaten to hold defense organizations back from achieving their mission outcomes. That is why it is so important for the Pentagon to overcome these obstacles. Here are three ways it can do so and develop an automation first culture.
Create a framework for automation across agencies
The Pentagon is taking an aggressive cross-agency approach to all-domain dominance through efforts focused on collaboration and information sharing, as evidenced by Joint All-Domain Command and Control (JADC2). The ambitious concept is all about connecting endpoints and data from across the breadth of the military and analyzing the data for better intelligence.
Automation has the potential to play a significant role in enabling JADC2 to reach its potential. The concept will require automated systems to quickly process and analyze massive amounts of data and create near real-time actionable information and recommendations. This can lead to better predictive analysis — the ability to anticipate the likelihood of an action based on historical data. Predictive analytics can also be powered by automation.
A total approach to aligning automation requirements with a program of the magnitude of JADC2 mission requirements must come from a strict emphasis on implementing a clear methodology that outlines procedures, processes and tools required for automation. The Pentagon must establish a framework that includes these components, as well as information on proper governance procedures, set policies, scope (depending on the needs of the agency) and expected capability outcomes.
Establishing such a framework that can easily be applied across agencies can:
- Ensure agencies adhere to the same automation principles and practices;
- Solve the problem of each agency pursuing their automation programs and goals, eliminating siloed “service-by-service” practices that too often crop up within government agencies;
- Supply agencies with a common set of priorities and support efforts like JADC2 by enabling the military to share and process information more quickly and easily without the need for human intervention.
Look to industry for inspiration
Although the concept of developing a framework is familiar to the Pentagon, there may still be some resistance to the idea of creating a framework for automation. Some of this may stem from past attempts to unsuccessfully implement automation, but those attempts may have been hampered by a lack of tools which have only recently become abundant.
For example, agencies that may have tried to automate before the advent of artificial intelligence, machine learning and other commonplace automation-powering technologies may have gotten burned by legacy technologies that were not yet up to the job. Like a person who has touched a hot stove once, they may be hesitant to try again.
But other industries have since successfully adopted and embraced automation. Consider the automobile industry, which uses automation to power decision making and actions by AI-driven autonomous vehicles. Or retail, which uses automation to provide shoppers a touchless shopping experience, including purchase recommendations. These industries have proven that automation is not something mysterious to be feared. It is not even really innovative anymore. For them, it is simply part of doing business while also creating a competitive advantage.
The Pentagon should adopt the same mindset and bake automation into its decision, funding and application development processes. Move it to the left, similar to how cybersecurity has shifted to the beginning of every conversation and embedded into development so that developers can automate and simplify their processes and better predict outcomes.
More importantly — and more ambitiously — defense agencies need to strongly consider building automation into the fabric of everything they do from an IT service management perspective. Soldiers need services delivered quickly and with great agility and flexibility. The only way to do this is through automation.
Improving the strategic process of delivering IT services requires mapping those services, creating blue-chip priorities, force or branch imperatives, and capabilities architectures. An automate everything approach can help the Pentagon achieve these objectives, and provide soldiers with the technologies they need for split-second decision-making.
Automation has the ability to transform the DoD, but to do so it must become a mission-critical priority on the same level as cybersecurity and other top initiatives. The time to do that is now.
Travis Steele is a senior strategic solutions architect for Red Hat Inc.