DOE Spotlight: Artificial Intelligence

The U.S. Department of Energy (DOE) and its National Laboratories maintain an innovation ecosystem to fulfill their missions in science, engineering, and national security. This ecosystem is expanding and improving AI systems, tools, methods, algorithms, and applications.

Critical Need for Artificial Intelligence

Artificial intelligence (AI) describes the capability of machines to rapidly learn from large data sets, solve problems, and continuously adapt to new data without human intervention. These machine systems leverage computer science, data science, and mathematics to deliver insight at data rates and scales incomprehensible to humans. AI functions include image and natural language processing, sensor networks, predictive planning, and decision support.

AI research, development, and demonstration (RD&D) has grown rapidly over the last 20 years, largely driven by advances in processors and big data. AI is now used in nearly every part of the economy to transform data stores into useful knowledge. Examples include targeted ads, face recognition, digital assistants, and self-driving vehicles. Global spending on AI systems is projected to reach $98 billion in 2023, up from the $37.5 billion forecast for 2019.

Top AI Industries Based on Projected 5 Year Growth Rates (2018-2023)

Media (33.7%)
Federal/Central Government (33.6%)
Resource Industries (32.8%)
Education (32.2%)
Personal and Consumer Services (32.0%)
Others (27.9%)

Source: International Data Corporation

DOE's Artificial Intelligence and Technology Office

To accelerate AI innovation and partnerships, DOE established the Artificial Intelligence and Technology Office (AITO). AITO serves as DOE's hub for coordinating the agency's efforts as a world-leading enterprise in scientific and technological discovery and accelerates the development, delivery, and adoption of AI.

National Strategies to Advance AI

DOE focuses its efforts on early-stage research to advance technologies and develop the next-generation computing capabilities, infrastructure, and tools that the nation needs but industry is unlikely to develop on its own.

Eight Strategic Priorities

Strategy 1: Make long-term investments in AI research Prioritize investments in the next generation of AI that will drive discovery and insight and enable the United States to remain a world leader in AI.

Strategy 2: Develop effective methods for human-AI collaboration Increase understanding of how to create AI systems that effectively complement and augment human capabilities.

Strategy 3: Understand and address the ethical, legal, and societal implications of AI Research AI systems that incorporate ethical, legal, and societal concerns through technical mechanisms.

Strategy 4: Ensure the safety and security of AI systems Advance knowledge of how to design AI systems that are reliable, dependable, safe, and trustworthy.

Strategy 5: Develop shared public datasets and environments for AI training and testing Develop and enable access to high-quality datasets and environments, as well as to testing and training resources.

Strategy 6: Measure and evaluate AI technologies through standards and benchmarks Develop a broad spectrum of evaluative techniques for AI, including technical standards and benchmarks.

Strategy 7: Better understand the national AI R&D workforce needs Improve opportunities for R&D workforce development to strategically foster an AI-ready workforce.

Strategy 8: Expand public-private partnerships to accelerate advances in AI Promote opportunities for sustained investment in AI R&D and for transitioning advances into practical capabilities.

DOE Develops AI Across Multiple Sectors

AI research at DOE is enabling solutions to critical issues in many economic sectors. Leading AI applications include grid optimization for renewable energy sources, transportation network efficiency, quality healthcare access, and advanced manufacturing processes.

Examples of DOE Program Office Funding Activities

AI Use in Science (Office of Science) The Office of Science announced $13 million in funding to develop AI as a tool for scientific investigation and prediction.

AI as a Tool to Improve Grid Resiliency (Office of Electricity) The Office of Electricity announced $7 million in federal funding for several AI-related projects to support faster grid analytics and modeling.

Advancing the State of the Art through Strategic Investments in AI (ARPA-E) ARPA-E announced up to $20 million for projects as part of the DIFFERENTIATE program, applying AI and machine learning to energy applications.

AI Applications by Sector

Grid Automation and Optimization

AI will facilitate grid modernization through autonomous systems optimization. The long-term goal is a fully Autonomous Energy Grid (AEG) that is:

Secure
Resilient
Scalable
Reliable
Affordable
Interoperable across devices in real-time
Robust
Flexible

Renewable Energy

AI can help to increase the generation and use of renewable energy in many ways:

Optimize dispatch of distributed energy resources
Improve forecasts of resource availability
Accelerate the discovery of new materials and technology

Geothermal: DOE research is applying AI to the exploration and production of geothermal resources. Successful implementation of machine learning methods could facilitate the discovery of geothermal wells, increase drilling accuracy, and reduce costs.

Biomass: Scientists at Idaho National Laboratory are using AI analysis of biorefinery processing data to guide operational adjustments that maximize output while mitigating system damage.

Transportation Systems

DOE and its National Laboratories use AI and machine learning to better understand and optimize transportation systems and urban mobility. These powerful tools are helping to improve safety and mobility while reducing congestion and energy use.

Benefits of Automation:

Reduce the 94% of accidents caused by human error (associated with 37,133 deaths in 2017)
Save the $242 billion in lost economic activity attributed to car crashes
Restore up to 50 minutes per day to the average driver
Provide 2 million people with access to jobs or career opportunities

Healthcare Applications

Advances in AI are unlocking new applications and approaches to healthcare. By applying machine learning, DOE's National Labs are improving the interpretation of medical and biological data and enabling advancements in diagnostics, drug discovery, and treatment.

CANDLE (Cancer Distributed Learning Environment): A partnership between DOE and the National Cancer Institute developing an open-source software platform using deep learning methodologies to explore cancer causes, treatment, and methods to improve patient outcomes.

Materials Discovery and Manufacturing

DOE National Laboratories are leveraging AI to advance the state of manufacturing and materials research:

Designing functional materials with tunable properties
Improving the efficiency of calculating complex chemistries
Accelerating the discovery of metallic glasses (200x faster than traditional methods)
Speeding the development of new catalysts
Screening new materials for solar cells

DOE Leverages Unique Capabilities for AI

National Laboratories with Core Capabilities in AI R&D

Argonne National Laboratory (ANL)
Brookhaven National Laboratory (BNL)
Lawrence Berkeley National Laboratory (LBNL)
Lawrence Livermore National Laboratory (LLNL)
Los Alamos National Laboratory (LANL)
National Energy Technology Laboratory (NETL)
Oak Ridge National Laboratory (ORNL)
Pacific Northwest National Laboratory (PNNL)
Sandia National Laboratories (SNL)

DOE Supercomputing Capabilities

DOE hosts four of the ten fastest supercomputers in the world:

Summit (ORNL): Currently the fastest computer, performing 200,000 trillion calculations per second (200 petaflops). Has enabled scientists to apply advanced machine learning to multiple topics.

Sierra (LLNL): At 125 petaflops, the world's second most powerful computer. Allows NNSA to run simulations on nuclear weapons in lieu of underground testing.

Trinity (SNL, LANL): Capable of operating at 41 Petaflops for maintaining a safe, reliable, and secure national nuclear stockpile.

Next-Generation Computing: Exascale

Exascale computing—able to do a quintillion computations per second—promises unprecedented breakthroughs in AI and machine learning.

2021 Scheduled Completions:

Aurora (Argonne National Laboratory)
Frontier (Oak Ridge National Laboratory)

2022 Expected:

El Capitan (Lawrence Livermore National Laboratory)

Key Partnerships

Exascale Computing Project (ECP): Driving the development of future exascale supercomputers and architectures through collaboration between DOE's Office of Science and National Nuclear Security Administration.

CANDLE: Collaboratively developed open-source software platform providing deep learning methodologies for cancer research.

HPC4 Energy Initiative: Facilitates partnerships between industry and national labs using DOE's high-performance computing resources for energy challenges in manufacturing, materials, and mobility.

ATOM (Accelerating Therapeutics for Opportunities in Medicine): Consortium to accelerate drug discovery using high-performance computing, biological data, and new biotechnologies.

AI Success Stories Highlighted

The document includes over 20 detailed success stories demonstrating AI applications across DOE and its National Laboratories, including:

Grid Optimization: Argonne's machine learning solving security constrained unit commitment 12x faster
Cancer Research: ORNL's MENNDL identifying cancer-fighting cells 16x faster
Weather Prediction: Advanced forecasting models using HPC and AI
Materials Discovery: SLAC discovering metallic glass 200x faster using machine learning
Healthcare: Deep learning for improved mammogram interpretation
Manufacturing: Machine learning reducing engineering design times from months to days
Neuromorphic Computing: Sandia's Whetstone enabling neural computing 100x more efficiently

This Spotlight document from the DOE Office of Technology Transitions demonstrates the breadth and depth of AI research across the Department of Energy's National Laboratory system and its critical importance to U.S. competitiveness and national security.