Twenty years ago, in the months and years following the events of September 11, 2001, agencies embarked on the important work of addressing information-sharing challenges by re-examining what they were collecting, why they were collecting it, and how they were holding it. Strategic decisions on information use and the governing decisions for access became central themes in information-sharing discussions. This work resulted in increased transparency within and across agencies on current information-sharing agreements, and most importantly drove a shift in the culture of “need to know” to “responsibility to share.” Addressing the information-sharing challenges across government also meant addressing the safeguarding responsibility for information. The combined information sharing and safeguarding demands pushed policy and governance processes to expand and become more transparent, and challenged decision making within agencies.
As a result of a sustained efforts, much progress was made as agencies redesigned policies, processes, and technologies, and challenged existing culture surrounding information sharing. Agencies put in place programs, architectures and decision-making bodies to overcome obstacles. This worked to enhance and improve access to existing data as well as determining what new data was needed to respond to threats across the broad Homeland Security community. Compared to the environment before the events of 9/11, the level of information and data that is shared within agencies, across agencies and with partners across all levels of jurisdictions is remarkable. Significant progress had been made; in February 2017, the Government Accountability Office (GAO) removed “Establishing Effective Mechanisms for Sharing and Managing Terrorism-Related Information to Protect the Homeland” from the high-risk list.
The pace of data generation, collection and dissemination is and will continue to increase exponentially, based on advances in data transfer capacity, mobile network adoption of 5G, lowered cost of data storage available in cloud computing, and a growing digitally native population. This has resulted in significant increases in the size of data collected and the speed at which is collected. For example, as smart phones have evolved, the size of the photos and videos captured has increased at least tenfold in the past five years. Agencies may be collecting the same types of data they were 5 or 10 years ago, but now need help to address increases in data size and ability to accept faster streams of data. So much so the explosion of data has exceeded the collective ability of the agencies involved in the Homeland Security community to understand or find the most relevant data to make analytical judgments for their mission. This challenge is just beginning, as new and innovative technologies emerge that are ripe for adoption by the Homeland Security community, including the following:
- Development of small, self-powered drones, with long-range capacity and ability to transmit data at very high speeds, can augment border security and port protection missions.
- Bio-sensing advances and approaches (pulse, blood pressure, blood gases, and blood sugar) developed for the personal health market may be leveraged to optimize search-and-rescue teams during recovery operations. These can be used to protect first responders during response and could potentially be used in select insider threat monitoring use cases for highly sensitive data.
- Light detection and radar/infrared data capture at increased speed and granularity can be used for increased accuracy of face-based biometric matching, assisting in threat-based screening, entry/exit programs, and physical security protection.
- Smart home internet-connected devices – an application of the IOT – can be used today to assist an aging population and could also be used to increase protection of the workforce in high-threat situations, enable faster situational awareness and optimize the protection of VIPs, dignitaries, and others at National Special Security Events.
- Robotic visual inspection devices, used in automotive and aeronautic manufacturing for defect identification, can be used for predictive maintenance, thereby allowing a reduction in operations costs and the optimization of funding to recapitalize aging fleet assets and infrastructure.
These new technologies create tremendous opportunities for agencies to drive significant innovation, insight, and decision-making in support of their mission. However, they must first tackle challenges in leveraging data now that, if not addressed, will become obstacles in optimizing the explosion of data in the future. Data that is shared and safeguarded still needs to be understood, both by machines and humans, before it can be used to make decisions. It routinely takes significant effort by both sending and receiving parties to make data useful across a multitude of data systems. It is common to hear that up to 80 percent of a data scientist’s time is spent formatting, ingesting, processing, integrating, correlating, aggregating, and managing data before they can use it to perform analyses that realize the value of data in their holdings.
“Agencies may be collecting the same types of data they were 5 or 10 years ago, but now need help to address increases in data size and ability to accept faster streams of data”
The recent establishment of a Chief Data Officer (CDO) in the federal government, driven by the Public Law No: 115-435 (01/14/2019) Foundations for Evidence-Based Policymaking Act of 2018, has resulted in a concerted effort within agencies to continue to push innovation to drive better, faster and less labor-intensive ways to analyze data, connect disparate data sets, apply context to data, infer meaning from data, and ultimately allow data science staff more time to build analytical products in support of agency missions.
It is an exciting time to be an agency CDO and the challenge may seem daunting. The demand for mission impact for a CDO is high and there is so much to do. Agency CDOs have a strategy to continue to make it easier, faster, and cheaper to use data at the enterprise level. They are organizing data management and governance structures to improve data effectiveness, reduce program costs, and unify the use of data across enterprise systems. Many are accelerating the adoption of mature data initiatives; building a coherent data ecosystem; acquiring the appropriate tools; reshaping the workforce; adapting new workflow processes; and changing the culture.
On the path ahead lies the promise that Intelligent Automation will transform an agency’s ability to move beyond human capacity, in finding value in enormous amounts of data. Intelligent Automation is changing the way business is conducted in every sector of our economy. Agency CDOs are well-positioned to make it happen more efficiently if they leverage relationships across their “C” suite partners and listen to the business stakeholder’s priorities.
In preparation for the broader use and adoption of Intelligent Automation, CDOs should also consider focusing on the following:
- Extend data governance structures to address ethical use of artificial intelligence.
CDOs are well positioned to help shape policies and authorities that govern how agencies deploy Intelligent Automation systems to help protect privacy, civil rights, and civil liberties, and ensure ethical use. Intelligent Automation systems present challenges to public transparency, specifically the ability to predict and explain the behavior of autonomous systems. In addition, the legal limitations on the use of personal information may limit the ability to feed Personally Identifiable Information (PII) into a system for the explicit purpose of making the system smarter.
Because algorithmically informed decisions have the potential for significant social impact, they must be designed and implemented in publicly accountable ways, such as an obligation to report, explain and justify specific decisions, as well as mitigate negative impacts and potential harm. Ensuring these protections are robust requires oversight for the implementation of ethical algorithms that do not create discriminatory or unjust impacts when making comparisons across different demographics or affected communities and individuals. CDOs, partnering with oversight bodies, can bring immense value with strong relationships already established for data governance by adopting best practices to review algorithms and the training data used for implicit and explicit bias, especially when produced or contained in proprietary solutions. Growing talent in this area may mean working with universities on internships to cultivate a potential talent pipeline. Moreover, building a sustained program of investment in the creation of high-quality training sets for priorities, designed for ethical use, may pay dividends for years to come.
- Use their new position to help advance digital evolution, in partnership with the CIO and CTO.
CDOs, CIOs and CTOs can help drive the right investments needed for their success to position agencies to reap the full benefits of Intelligent Automation. CDOs are reliant on their peers to make progress. The CDO to CIO relationship is critical, as CDOs need a modern infrastructure on which to build their capabilities. Prioritizing investments evenly between the CIO and CDO can ensure agencies are prepared to address emerging issues years later. For example, the U.S. Digital Service, founded by President Obama in August 2014, brought together the best engineering, design and government talent to change our government’s approach to technology. After a few years of success by early adopters of the U.S. Digital Service, many agencies recognized the importance of embracing the digital revolution and began to increase investments in adopting these capabilities. In 2020, when the pandemic emerged, agencies that implemented a solid foundation of digital capabilities through cloud adoption found themselves able to pivot quickly, seamlessly transition to telework, and rapidly respond to major policy initiatives.
It is time to look at investments across the Intelligent Automation spectrum in the same manner. CDOs can measure their success in the speed at which their agency can ingest new sources of data securely and create value for the mission. CDOs should participate in annual investment reviews conducted by the CIO to assist in asking questions like “Does this investment add capabilities needed to adopt advanced technologies such as artificial intelligence?” or “Is it the right investment now, or are the funds better spent on modernization infrastructure?”
CDOs in agencies with limited digital, data, or people maturity may be struggling with technology modernization, have significant backlogs of technical work, be slow to implement modern development methods and may be struggling to adopt cloud technologies. Implementations of RPA will arise as short-term fixes where systems are not integrated, data does not flow from one to another, and the time or cost to modify these systems outweighs the benefit. Since adoption of RPA requires minimal training, adoption can spread very quickly. Unchecked, this may hide systemic issues and add to the backlog of technical work.
Working with a CTO assessing investments for the agency in this situation may mean advocating for cloud computing, network modernization or maturing agile competencies, all of which will contribute to their success. CDOs can monitor where RPA is being adopted quickly to find hot spots, then help capture cost/benefit data to drive the decision of when to make investments to overcome the need for RPA.
CDOs in agencies with a moderate to high level of digital, data, or people maturity may be deploying types of Intelligent Automation at all levels of complexity. CDOs can work with the CTO and CIO to help balance investments across a set of maturing initiatives such as:
- Advancing cloud migrations to ensure full visibility of monitoring applications, combined network and security operations, accelerate processes to achieve Authority to Operate, and continually optimize consumption.
- Modernizing the network or pushing zero trust approaches ensuring availability of secure computing to the mission edge.
- Advancing data infrastructure and platforms to optimize data collected from cyber sources for advance threat analytics that will arise from new applications of Intelligent Automation.
- Preparing for next-level technologies by working with the research entities to understand potential impacts of their work, oversight requirements and speed required for data ingestion.
CDOs have a daunting task ahead, and success may only result by using their “seat at the table,” deepening partnerships across the “C= suite”, and taking an enterprise view to ensure a coordinated investment in a rock-solid foundation. Agencies with CDOs ready for these challenges will push innovative ways to leverage data, harness the power of unique data collections and provide decision advantage at unmatched machine speeds well into the future.
