Bo Yang during the interview
The United States’ power infrastructure is at a pivotal point in its history. Artificial intelligence (AI) data centers, expanding electric vehicle (EV) charging networks, reshored manufacturing, and more, are accelerating the electricity demand beyond the flat growth rates the grid was designed around. As demand climbs, the need for dramatic infrastructure updates is intensifying. This is not a simple problem to solve. Careful planning, modeling and simulation are required to determine how to update these complex system designs without creating undue risk.
At the center of this research for Hitachi is Bo Yang, Vice President and Head of the Energy Solution Lab at Hitachi America R&D. Trained as a power systems engineer, Yang began her career conducting forward-looking national lab research on large-scale renewables, storage, and high-performance computing. From there, she moved into engineering, consulting on daily utility engagement at GE and Siemens. Subsequently, she moved into product development and deployment in the power electronics industry. Now, she leads Hitachi’s efforts to build practical, AI-enabled grid solutions.
“Over the years, my experience has provided me with a lot of training,” shared Yang, “including how to define a product, how to leverage advanced digital technologies such as AI, and how to make a real impact in today’s energy world.”
This progression from research to hands-on utility problem-solving has shaped how Yang and the Hitachi team approach solving contemporary power infrastructure challenges. “Hitachi is where I get to work with real-world problems,” emphasized Yang, “leveraging the latest advanced AI and digital technology while being very practical and generating the right impact.”
Bo Yang during the interview
Up until the past few years, the U.S. electrical grid experienced relatively flat demand, seeing only 1-2% annual load growth. This remarkable level of stability served as the baseline around which utility companies built their systems and infrastructure. However, this foundation has all but been upended by the explosive rise of AI.
“We see the rapid booming of the AI industry increasing the number of data center projects being launched across the country,” noted Yang. The surge in large-scale data center construction and expansion drives an exponential increase in the amount of computational power and the electricity necessary to support it. At the same time, the expansion of EV charging networks, renewed manufacturing facility growth are placing even more high-intensity load on the power grid.
Meeting this rapidly increasing demand requires that energy generation sources—new, renewable, and conventional—be brought online quickly. Unfortunately, this tends to be anything but a fast process. “New power generators cannot be plugged into the grid overnight,” Yang highlighted. “We must follow a very strict review and regulation process.”
Every new energy project in the United States must undergo a rigorous, federally regulated interconnection review to ensure that it will not introduce instability or risk overloading the system. Each study involves extensive system modeling and simulation, contingency analysis, and stakeholder reviews, which inevitably slow the pace of integration. What’s more, most utility organizations still rely on simulation tools from decades ago, which can take days to run a single high-complexity scenario and significantly delay these projects.
The longer these projects take to approve and deploy, the greater the risk of grid failure. Modern load increases are straining the system’s capacity reserves, resources that are meant to buffer against outages during storms, heat waves, or other unexpected equipment failures. If these resources are exhausted now, they’ll leave the grid more vulnerable to the consequences of extreme, unanticipated events. This widening reliability gap, spurred by increasing demand and slow resource additions, is not sustainable for our world’s growing digital workloads.
Bo Yang during the interview
Modernizing the process of grid planning to account for these challenges requires more than just incremental improvements. It necessitates a greater shift in the interconnection review process, one that brings more dynamic simulation, modeling, and decision-making capabilities to the forefront. This is why Yang and the Hitachi team are integrating AI solutions and GPU-accelerated computing into grid planning and analysis.
The goal of leveraging these solutions is to strengthen engineering fundamentals, not replace them. “We want to interweave advanced AI techniques with the modeling-based approach, so that the AI model can comply with all the safety constraints that the power industry is going to mandate,” Yang explained.
One significant benefit of this integrated approach is acceleration. While traditional simulation tools can take over 40 hours to complete a single study, AI and GPU acceleration can reduce this time significantly. This takes modeling and simulation from a multi-day process to an hours-long project, allowing planners to analyze a larger number of scenarios and better understand interconnection risks. “With much shorter simulation time, they can potentially study ten times more scenarios at a much more granular level,” explained Yang.
Beyond increased speed, AI and accelerated computing solutions ultimately streamline the entire study and review lifecycle. According to Yang, they can “automate the overall study flow, minimizing a lot of steps that may introduce error.” Instead of relying on engineers to move data between formats, consolidate inputs, and validate model integrity, AI can create a pipeline that’s free of human error and enables repeatable research at scale.
Crucially, Hitachi approaches this integration not as a technology-first exercise, but as a people-centered transformation. Yang and her team emphasize the importance of building with users, not merely for them, to ensure that AI-enhanced interfaces, analytics, and automations align with how planners and operators are actually making critical grid decisions.
Hitachi advances this pipeline through co-creation that combines the OT expertise of Hitachi Energy, digital and data capabilities from Hitachi Digital, and input from utility partners, policymakers, and research institutions. By piloting purpose-built solutions in customer environments and refining them through real data and operational conditions, the team can deploy practical, compliant, and future-ready solutions.
Bo Yang during the interview
While AI integration has a noteworthy impact on long-range interconnection planning, its impact on real-time power grid operations may be even more significant. Today’s control centers must manage rapidly changing grid operations with fluctuating loads, introducing a level of volatility that was not previously a concern. “The grid is changing in real time,” noted Yang. “The top priority is to make sure there’s no glitch that will cause this grid to fail.”
This volatility places an increased premium on visibility, forecasting, and rapid response capabilities—all of which can be bolstered by AI solutions. AI integration can arm grid operators with:
AI models can quickly analyze various weather patterns, price signals, and demand trends at more fine-tuned temporal and spatial scales, helping operators anticipate stress points before they materialize. This improved forecasting can help utility organizations avoid sudden price spikes and better manage expected surges.
According to Yang, the enhanced forecasting will allow them to “operate the wholesale market so that they can minimize the total energy price for customers,” making more proactive, cost-effective decisions in a generally uncertain system.
Traditional asset management approaches, such as vegetation control, often rely on operator intuition or fixed maintenance schedules. With AI solutions, teams can analyze imagery, identify risks, and more effectively optimize maintenance routes to avoid unexpected obstacles.
“AI can give you much more information,” claimed Yang, “helping you optimize the routing and where to go first.” When extreme weather does hit, AI can help support outage detection and notification, restoration routing, and repair prioritization, making recovery efforts faster and more coordinated.
Control room operators must make countless difficult decisions while navigating the complexity of the modern grid. This is only made more difficult by the increasingly distributed resources, dynamic load patterns, and interdependent systems present in today’s systems.
Operators need tools that can synthesize immense amounts of data from across various sources and highlight key decision-making insights. By surfacing risks earlier and offering contextually informed guidance, AI can help optimize grid operators’ decision making as a critical support tool, not a replacement for human judgment.
Bo Yang during the interview
“We want to make sure that our technology is solving the right industry problem and making the right impact,” shared Yang. The goal of her team is not just to achieve faster simulations, but to empower better decisions. By pairing AI solutions with proven engineering methods, Hitachi gives utilities organizations the time and clarity to plan more confidently and operate more proactively. They’re preparing grid operators to respond to future volatility by anticipating stress, streamlining restoration, and keeping energy reliable and affordable no matter the scenario.
Related Link
