How High-Performance Computing Can Strengthen the U.S. Government

Digital abstract Art/

More computing power means more mission delivery.

The U.S. government currently uses high-performance computing technology in a variety of ways and experts expect increased investment in artificial intelligence for use within the federal space. This article will explore the ways in which the U.S. government is currently adopting HPC and AI among its agencies, where agencies are already seeing results, and what the future will look like.

The Securities and Exchange Commission, for example, is currently implementing AI by way of agency enforcement activities and has developed four algorithmic tools that are able to detect various forms of fraud. The Corporate Issuers Risk Assessment detects fraud in accounting and financial reporting. The dashboard features around 200 metrics that can find anomalies in the financial reporting of more than 7,000 corporate issuers of securities. The Advanced Relational Trading Enforcement Metrics Investigation System, or ARTEMIS, and the Abnormal Trading and Link Analysis System, aka ATLAS, look for suspicious trading activity. ARTEMIS specifically hunts for potential serial insider trading offenders while ATLAS analyzes for first-time offenders. Finally, the Form ADV Fraud Predictor helps SEC staff predict which financial services professional may be violating federal securities laws.

Military aviation maintenance is critical to the success of countless missions, training and programs. Commercial AI applications are being used by military aviation maintenance engineers to help predict when an aircraft needs repairs or when aircraft parts need to be replaced. In doing so, overall efficiency will be improved and more aircraft will be available for missions. The use of AI can also help provide new ways to design and develop lighter, more efficient aircraft, which will help to reduce costs in the long term. 

Law enforcement is currently using AI in a number of ways including gunshot detection, crime forecasting, DNA analysis, and public safety video and image analysis. Many of these key areas are labor-intensive and require a subject matter expert to garner results. Once an AI tool is trained, it can run simulations and arrive at insights and predictions quickly. Law enforcement can use these AI derived insights to better allocate resources in an effort to more efficiently prevent and solve crime. 

AI has also helped in managing the spread of COVID-19 by using Video AI to conduct temperature scans, detect masks and handwashing, monitor contacts and social distancing, and perform contact tracing.  By collecting and analyzing this data, AI can contact trace infected people to identify which employees to quarantine.  This targeted selection can save companies millions in closure costs. By analyzing all of the data and using trending algorithms, it can cross correlate many dimensions to provide better accuracy. Video AI has advanced to the point that it can even predict future outcomes before they happen in better than real-time scenarios. The Transportation Security Administration, Homeland Security Department, Veterans Affairs Department, and National Institutes of Health are all agencies that have the potential for strong use cases with this technology.

The Future of AI in the U.S. Government

Before the federal government can implement AI to the fullest extent, they need to recruit talent. Government agencies face challenges accessing and recruiting talent with the necessary technical skills to implement and further develop AI technology. One major hurdle is that civil service laws cap government salaries far below competitive salaries offered by technology companies (source: ACUS AI Report). To ease this burden, the Office of Personnel Management have taken steps to create new classifications for technical positions and establish a governmentwide “direct hire” appointing authority for a number of STEM positions (source: Office of Personnel Management).

With humans and AI working harmoniously together, the possibilities for a stronger and more streamlined public sector are endless. The impact will be unprecedented as more agencies continue to implement highly sophisticated applications that will help with efficiency, accuracy, and workflow.

Sid Mair is president of Penguin Computing.