Skip to content
Fast Company Executive Board

The Fast Company Executive Board is a private, fee-based network of influential leaders, experts, executives, and entrepreneurs who share their insights with our audience.

Why you need scenario planning in 4D

The next big thing in AI will transform infrastructure management.

[Photo: Getty Images]

Profile photo of Todd Slind

Vice President of Technology · TRC Companies

Over the past decade, digital twins have become powerful tools for organizations with complex infrastructure. These virtual models of the physical world, combined with SCADA signals, give operational teams an unprecedented level of visibility into their most critical systems, helping them confidently answer what is happening with their infrastructure right now. 

But showing that is only part of what operational teams need. Organizations also need to know if something else may happen to their infrastructure.

Digital twins share what is happening right now, but they can’t answer questions about hypothetical future scenarios. Yes, digital twins have some capabilities for forward-looking activities like predictive maintenance, but that is not robust enough for the scenario planning organizations need to do around threats to their operations such as:

  • What if an ice storm hits our critical infrastructure in this region?
  • What if there is an outage with this system that lasts hours or days?
  • What if the frequency and intensity of these types of extreme weather events produce compounding impacts to our infrastructure?
  • What if there is a supply chain disruption for critical components over the next few years?
  • What are the areas of highest grid stress?
  • What are the risks to our operations if X, Y, or Z happens?

In a world where change is accelerating in ways that make risk management more difficult, this kind of scenario planning is vital for organizations. A different technology is needed for it, and that is where the next big thing in corporate AI comes in: world models.

WORLD MODELS

World models are the result of a major effort among experts to apply AI technology in ways more attuned to business needs, versus tools that were originally designed for more abstract artificial general intelligence goals. It is AI for the physical world, the reason it is such a compelling fit for managing infrastructure.

World models are not a new concept. They just took a few decades for technology to catch up and make it a reality. In the mid-20th century, Kenneth Craik conceptualized the idea of 4D world models that would utilize AI to interlink human cognition and technology. Today, that vision is becoming a reality through billions of dollars in investment by the world’s largest tech companies.

A prominent example of this is Fei-Fei Li’s World Labs, an AI company focused on building spatial intelligence models. The company recently secured more than a billion dollars in funding from backers such as NVIDIA, Autodesk, AMD, Emerson Collective, and Fidelity to accelerate its world model research and development (and as of late September, agreed to be acquired by AMD). They have already brought a robust early version to market that gives businesses the ability to start putting this technology to use. For world models, this is the equivalent of when ChatGPT was first released, allowing companies to begin exploring how to put it to work.

World models create 4D digital models of physical environments by aggregating, integrating, and analyzing organizational data into a dynamic view of physical environments. And world models are constantly updated in real time as new data is collected or as blind spots are eliminated using techniques like 3D computer vision that I discussed previously.

For those of you familiar with Star Trek, world models are the equivalent of the holodeck, enabling users to create hyper-realistic virtual worlds that let you play out scenarios, conduct training, and plan strategies based on those realistic simulations. But instead of simulating battles with Klingons or training for high-stakes trips to new planets, world models enable organizations to think through and prepare for what-if scenarios—and harness the power of AI to proactively solve them.

WORLD MODELS IN ACTION

What would this look like in action? Utilities could use world models to realistically model skyrocketing load growth. Or the impact of a mega-storm. Or the fast-changing risk of wildfire. And utilities can use the insights from world model-driven scenario planning to make smart decisions about where to upgrade infrastructure, where to move infrastructure underground, where to deploy storm response teams, and more.

Similarly, other industries can use these models to prepare their infrastructure and operations for the impact of climate change, supply chain disruptions, natural disasters, security vulnerabilities, large events (e.g., World Cup, Olympics), and more.

For industries with large-scale infrastructure at the heart of their operations, world models are a powerful new tool for protecting operations in an era when the pace of change is accelerating and operational risk is increasing in significant ways.

Todd Slind is VP of technology at TRC, a member of WSP.