By mid-century, the planet will be home to nearly ten billion people. To ensure that everyone has secure access to energy, we need to accelerate the transition to clean and renewable sources. We must do so while protecting the earth’s ecosystems and ensuring equitable access to safe and nutritious food. This means that while we drive the potentially space-demanding and mineral-intensive decarbonization of society and industry, we must transition to green shipping, renewable energy sources, and sustainable food systems – and protect biodiversity through the careful management of both land and ocean.
Although there is no single solution to these diverse yet interconnected challenges, it is becoming increasingly clear that various artificial intelligence (AI) and machine learning (ML) technologies can contribute to many of their solutions. Here are just a few ways in which the data-processing abilities of AI and ML can address issues ranging from climate change and biodiversity loss to food insecurity and inadequate health services:
They can help limit and mitigate climate change by optimizing energy consumption, supporting the integration of renewables into smart power grids, facilitating resource-efficient precision agriculture, and reducing emissions from ship operations.
They can aid in land- and ocean-based species monitoring and conservation efforts through image recognition, sensor data analysis, and predictive modelling.
They can improve diagnostics through medical imaging analysis, enhance patient care through personalized treatment plans, assist in drug discovery and development, and streamline healthcare operations for greater efficiency and cost-effectiveness.
However, we should not have unrealistic expectations. AI will not automatically bring about the desired changes and transitions, and it will not necessarily do so in a safe, transparent, ethical, and non-discriminatory way. Although AI technologies are being developed at an unprecedented rate, they will not be deployed successfully unless affected stakeholders feel confident that they are safe and deliver the intended value. Hence, to accelerate the safe adoption of AI solutions, we need regulations and assurance frameworks that build the necessary trust.
In 2023, DNV completed several research and development projects that generated insights and methods to help stakeholders verify and demonstrate the safety and quality of their digital technologies. Among other things, we published a recommended practice (RP) on how to assure the trustworthiness of AI-enabled systems. This RP describes a comprehensive and concrete assurance approach based on a thorough understanding of the potential benefits and failure modes of AI-based systems. You can read more about this RP and other selected projects in this [publication title].
As we look towards 2024 and beyond, our commitment to the energy transition and trustworthy AI remains firm. Our research efforts will focus on hydrogen safety, ammonia as a ship fuel, and power management in systems with a growing share of renewables, among other things. We will continue to publish our annual Energy Transition Outlook report, as well as regional and sector-specific deep-dive reports.
In addition, we will intensify our efforts in biodiversity research to learn more about how human activities affect the environment and ecosystems. We will lead the way in exploring ways of generating high-quality synthetic data that can be used to train AI models, not least in healthcare applications. As we explore safe and sustainable alternatives to fossil fuels, we will extend our research into nuclear propulsion. Last but not least, we will step up our cybersecurity R&D efforts. As AI continues to change the cyber landscape, we will deepen our understanding of how AI can be used by cybercriminals, how it can be used for cyberdefence purposes, how AI itself can become a target of cyberattacks, and how to build resilient system architectures that can withstand new and evolving cyberthreats.
Our research will be deep and specialized to generate actionable, sector-specific insights. However, it will also be broad and multidisciplinary enough to capture the global, complex, and entangled nature of the issues we aim to address. We believe that by working with industry partners, academia, and regulators, we can provide science-based insights and assurance frameworks that will unlock the potential of new and promising technologies.
I joined DNV and the GRD Healthcare Programme in August 2022. My previous experience includes a PhD in biomedical engineering and many years of industrial R&D in various industries and projects, such as 3D cameras and cardiac ultrasound. The projects provided me with hands-on experience with AI development, from concept to deployment.
