FORT HUACHUCA, Ariz., The renowned theoretical physicist, Stephen Hawking once said, “Computers will overtake humans with Artificial Intelligence (AI) within the next 100 years. When that happens, we need to make sure the computers have goals aligned with ours.” Likewise, what most frequently drives AI in this day and age is data science and according to Hawking this is a big part of cyberspace with its own set of difficulty. “In a dazzlingly complex world, you have to be able to discern the meaning in the mess. We are, in a figurative and literal sense, awash with what we call data.”
“What we’re only now fully realizing is two-fold: the sheer quantity of data in any given domain; and the tools we need to make use of the information encoded in it,” said Hawking.
As the U.S. Army continues to operate in a heavily contested cyber environment, the Network Enterprise Technology Command (NETCOM) is using data as a means to gain a strategic digital advantage over our cyberspace adversaries while synchronously applying data science to influence AI and gain foresight that achieves a dominant tactical advantage within any Joint Area of Operations (JOA) that requires data-centric network enterprise support.
Furthermore, cyber security analysts often refer to data as the weapon of choice when addressing cyber threats by both peer and state sponsored adversaries.
“From a cybersecurity standpoint, the volume and velocity of threat and vulnerability data make it very difficult for analysts unaided by data science to secure the network or recognize and respond to threats,” said Col. Eric S. Tollefson, NETCOM Director of Data Science.
Using data science to influence DoD’s inherent battlespace across all echelons of the Joint All Domain Operations (JADO) environment is not a new concept; thus, the Department of Defense (DoD) is currently unleashing data to advance the National Defense Strategy (NDS) and Digital Modernization (DM) initiatives. Presently, the DoD Data Strategy provides overarching vision, focus areas, guiding principles, essential capabilities and goals necessary to transform the Department into data-centric enterprise.
Supporting DoD’s data-centric enterprise initiative is the NETCOM Data Science Directorate (DSD) via the Army’s Data Plan.
“Although, the Army does not have a Data Science Strategy per-say, it does have the Army Data Plan and its corresponding HQDA EXORD 009-20, Army Data Plan Implementation in Support of Cloud Migration, as well as The Enterprise Data Analytics Strategy for Army Business: 2018-2022,” said Tollefson.
“As consumers of data, the DSD supports the Army Data Plan by developing data requirements essential to operationalize data science at NETCOM, which in turn helps the data management and governance bodies prioritize their efforts. With the addition of the NETCOM data management function to DSD’s portfolio, we now also directly support that plan by implementing data management at the NETCOM level,” added Tollefson.
According to Tollefson, the NETCOM DSD also directly supports other key Army and DoD initiatives focused on AI and Machine Learning (ML).
Nevertheless, as so happens with any inter-disciplinary field, data science remains a highly analytical and practical science that often has its share of challenges.
“For any given problem, we struggle with finding the necessary data (visibility), getting to the data (accessibility), knowing what’s in the data source (understandability), and having high quality data (trustworthiness),” says Tollefson.
“There are scores of examples we could point out. We spend a large portion of our time dealing with data-related challenges. However, that’s not unusual from what data scientists from elsewhere in the Army and DoD, as well as in industry, face; dealing with data issues is simply part of the process,” added Tollefson.
NETCOM’s DSD focuses its mission on providing integrated and advanced analytics capabilities in order to enable objective decision-making. In addition, to meet mission requirements and to maintain a technological edge over peer-to-peer adversaries, the NETCOM DSD maintains an emphasis on its Lines of Effort, which include providing data analytics products and services, maximizing use of enterprise data, developing partnerships, building capabilities and managing talent.
“We have also recently picked up the mission to lead data management across the command. We have begun work on implementing that capability; however, this is still a work in progress,” emphasized Tollefson.
Not only has the NETCOM DSD remained postured and engaged with providing the command data science expertise, the directorate has made great strides in collaborating with the Army, DoD and institutions of higher learning during COVID-19.
“Throughout COVID, DSD coordinated, collaborated and connected in a distributed manner due to our organizational structure,” added Tollefson.
“In addition to our main body at Ft Huachuca, we are operating small offices in three satellite locations across the country (Phoenix, Monterey, and Pittsburgh), which were stood up for the purpose of partnering with academia and industry and providing access to the highly qualified talent pools of those major metropolitan areas.”
“Moreover, in order to work together seamlessly, we had to implement tools and develop processes for collaboration that prepared us well for the COVID-19 environment. The tools implemented by DoD and the Army during COVID-19 led to even greater improvements in our ability to communicate. Data science is absolutely a team sport, and collaboration is critical,” explained Tollefson.
Nevertheless, keeping a team like the DSD working flawlessly through a global pandemic and operating with the parameters of a heavily contested cyber environment often takes a group of winners and over-achievers with the technical expertise to meet the challenges of sustaining a data science eco-system that not only supports the U.S. Amy but the DoD as well. So in order to maintain a dominating competitive posture, the team often competes in Army and DoD wide competitions designed to showcase the limitless talents of this unique NETCOM organization.
Although, officially established in 2017, the DSD team has hit the ground winning accolades in several competitions and recognition throughout the Army and DoD.
“DSD placed first in the Data Visualization category during the 1st Army Analyst Data Challenge (AADC) in 2019. That year, winners were selected from two categories – Data Visualization and Predictive Model. The visualization category judged how well teams represented their information and results for decision makers,” recalled Tollefson.
Equally, in 2020 the team also competed and placed in the AADC competition as well; however, balancing duties, COVID-19, remote operations and staying in-tune with real-world threats took a heavy toll on the team.
“Access to software and tools was and is the consistent problem. Each member did work on their local machine and it was not always easy, or possible to share work with one another due to accessibility issues,” said Lt. Col. Robert N. Collier, Senior Data Scientist at the NETCOM Phoenix Data Science Center (DSC).
“This is a very broad problem that is being earnestly worked on in the Army and elsewhere in the DoD. In short, it sounds easy, but one of the biggest challenges is that it is not easy / possible to either bring the data to the tools, or vice versa. Thus, if an individual can do it on their local machine, others in their group may not be able to do the same,” emphasized Collier.
With limited remote access to much needed software tools and external user constraints, the team remained unable to achieve the same success as the year prior.
“We came in third of over 30 teams. The top two teams did exceptional work, and deserve their due credit,” recalled Collier.
Other team members had this to say; “This was a very good finish, given the challenges with data and the data science environment along with ongoing project work. Last but not least getting to know and interact with like-minded team members was a significant plus that I really enjoyed,” said Tina Agarwal, Data Scientist at the NETCOM Pittsburgh, DSC.
According to the DSD AADC competition team lead, Maj. Nicholas L. Lee also echoed some of Collier’s thoughts regarding last year’s competition trials and tribulations.
“I think the greatest challenges these data science competitions bring to the table is the fact that they try to implement realistic constraints that exceed the teams to ability to collaborate across domains, locations, services, etc.,” evoked Lee.
“Remote collaboration is not an easy task, especially on complex problem sets. Fortunately for us, we are satellite based, and fully embraced this challenge by selecting members, who accustomed themselves to work from remote spots rather than operate close to the flagpole [brick and mortar locations] as they say in the Army,” recounted Lee.
Although challenged last year during the AADC competition, DSD team members found comradery as they explored innovative ways to problem-solve and articulate difficult project-sets during the contest.
“There are so many different, and creative ways to tackle a problem and being afforded the opportunity to participate in the 2020 AADC is a fantastic learning experience for me. Not to mention, the fact that we have a geographically dispersed team, having the chance to work alongside other Data Scientists that I don’t necessarily get to work with on a regular basis is another added benefit,” said Chris Fearnow, NETCOM Operations Research Analyst.
“I think NETCOM is a great organization, and the people work well together. And as far as the DSD is concerned, we are a tight knit group of individuals, who share knowledge, teach, mentor and progress in the data science field on a daily basis. Teamwork like this is often hard to find elsewhere,” added Lee.
Respectively, for a data science team like the NETCOM DSD that knows no bounds and has no limitations when it comes to its personnel and their diversified talents; there is no problem they cannot solve. And this assertiveness is frequently voiced quite persuasively by the team members themselves.
“I like how data science leverages skills in engineering, statistics and domain expertise as part of the problem that one is trying to solve. Though it’s difficult to excel in all the areas, one can find a niche and work towards more specialized roles like a Machine Learning Engineer, Data Engineer and Architect,” said Agarwal.
“I believe the challenge in data science revolves around data itself, especially when the data is messy and not available in the right format. However, I like the ability to research and collaborate as part of a small team and having the freedom and opportunity to overcome challenges in creative ways,” highlighted Agarwal.
As data competitions maintain skillsets and multi-talented personnel help integrate and extrapolate data management tactics, techniques, and procedures (TTP’s), the NETCOM DSD Team is remaining postured and ready to enable a data-driven enterprise that is fully capable of optimizing and transforming the Army’s portion of the Department of Defense Information Network (DoDIN-A). This stance also allows NETCOM to sustain uninhibited support of Army and DoD data science initiatives. This comportment then translates into a DSD team that is prepositioned to meet the new challenges of the information age.
“The Army is currently moving from the industrial age to the information age, leveraging data as a strategic asset. The work we do is fundamental to that transformation in the cyber domain,” said Tollefson.
“NETCOM leaders past and present have recognized that data science and data management are integral to how the Army will fight, and DSD is at the cutting edge of that. Furthermore, we are a self-learning organization, continually re-evaluating how we do things to ensure that we evolve as both NETCOM and the cyber domain evolve.”
“Additionally, we employ agile processes that focus on delivering value quickly and iteratively to the customer, providing greater operational flexibility and adaptability to a changing environment,” recapped Tollefson.
For more information on NETCOM, please visit: https://www.army.mil/netcom