Supercomputing and AI Collaborate to Simulate Everything from Earth to the Heart and Beyond

In the current industrial revolution, programming has shifted from tool creation to skill generation. This crucial step requires the best systems, data, and a secure environment to test them. “We need to simulate the world,” says Rev Lebaredian, Vice President of Omniverse and Simulation Technology at the tech giant Nvidia. This path has led to the creation of supercomputers and artificial intelligence (AI) systems with ambitious goals like creating a digital twin of the planet (the European plan Destination Earth). Additionally, it has fostered new ways to understand and observe the environment, leading to applications in everyday life that span conventional commerce to fashion, urban traffic management, building and water management, navigation, healthcare, and defense.

DestinE, as known in the European Commission program Destination Earth, aims to have an operational, highly accurate digital twin of Earth within five years for simulating and tracking natural phenomena, risks, and related human activities. The first supercomputer to join the project was LUMI, the most powerful in the European Union (EU), which HPE participates in.

“LUMI, until recently, was the largest supercomputer in Europe and remains the largest available to the climate research community in the EU. This allows for very high-resolution long-term simulations, including atmospheric, oceanic, and land surface models at a scale previously accessible only for short-term weather models. It can also be used for smaller workflows continuously or for urgent requests,” explains Utz-Uwe Haus, head of HPE HPC/AI EMEA Research Lab, a division of Hewlett Packard Labs.

The goal is not theoretical, rather it enhances our ability to better understand the specific aspects of an event for short-term decision-making, such as in disaster management, or climate change scenarios, including modeling the impacts of glaciers, sea ice, vegetation, and aerosol particles, among others, on Earth’s climate,” adds Haus.

These applications also include more local aspects, as specified by the researcher: “A high-resolution climate twin makes it feasible to predict very specific local effects that translate directly into social and economic impacts: average rainfall, drought periods, and flood scenarios within a city or region are critical parameters in agricultural land use and crop planning. Similarly, the economic viability of investments in renewable energy sources largely depends on average predictions of sunlight or wind strength for the decades of technological lifespan.”

However, the field of this convergence of supercomputing, artificial intelligence, and digital twins is not limited to virtual spaces. “They blur the line between the physical and the virtual. More than simulations, these virtual copies of reality incorporate real data to reshape how we design and build, allowing us to predict real-world outcomes with astonishing accuracy,” states the European Organization for Nuclear Research (CERN) following a workshop on the complexities of this technology and its applications.

They are key in high-energy physics experiments and for enhancing the operability of robots or for developing cooling and ventilation systems or optimizing production and ensuring equipment safety in manufacturing and industry. In healthcare, they aid in patient data analysis, surgical planning, and training. They play a fundamental role in design, testing, and simulation across aerospace, automotive, and aviation industries. They also significantly contribute to smart city initiatives, helping in urban planning and infrastructure management.

The Twin of a River

An example of these uses is the digital twin of the Eurovia de Guadalquivir, included in the Digitalization Plan of the Port of Seville, with over three million euros already invested and around twenty projects underway, according to the Port Authority president, Rafael Carmona. He explains that Guadaltwin, as the twin is called, “will integrate machine learning and artificial intelligence technologies to enhance the accuracy of predictions and decision-making.”

Similarly, the metropolitan area of Barcelona has a digital twin of its 164 municipalities to identify objectives for the next two decades and simulate urban, economic, knowledge, mobility, and housing scenarios.

Risk-Free Trials

In the healthcare sector, Steven Niederer, a biomedical engineer at the Alan Turing Institute and Imperial College London, works on detailed digital replicas of human hearts. They function the same way, in the same size and shape as real hearts, but their virtual existence allows for simulating surgeries without any danger. These models are already being used in clinical trials and patient care. They are a precursor to digital versions of the human body that will enable anticipating risks, trial interventions, and determining which treatments might work best.

“Treating arrhythmias such as persistent atrial fibrillation or deciding whether a patient would benefit from an implantable defibrillator to prevent sudden death remains, in many cases, a top clinical challenge. We need to develop new technologies to address these goals, and we have already taken the first steps: digital cardiac twins provide a model that allows not only visualization but also simulating heart behavior and anticipating responses to different treatments,” write Andreu Climent and María de la Salud Guillem from the Polytechnic University of Valencia in The Conversation.

In the defense sector, companies like Raytheon are working on digital twins of missile defense radars, air-launched cruise missiles, and aircraft.

And this trend extends to more mundane elements. The fashion brand H&M, as reported by BOF, has created digital replicas of 30 models to recreate them with AI in promotional campaigns. The company states that the rights to the twins will belong to their original models, but acknowledges the impact of this initiative on the teams involved in its campaigns, such as photographers, stylists, hairdressers, and makeup artists, among others.