Digital Twins, Geospatial AI Help Bridge the Physical World and Digital World

Meeting of the Minds

No one can see the future, but we can certainly try. Armed with new predictive capabilities such as digital twins and geospatial artificial intelligence (AI) – advancements that seem more at home in a sci-fi novel than in the hands of utilities – these new tools and technologies may make the ability to “see the future” more of a reality than we may think.

The ability to collect, manage, enrich, and analyze large amounts of data is driving organizations to not become “smarter,” but masters of knowledge. This new ability allows the tacit knowledge of experience and expertise to be converted to the explicit codified knowledge that gets work done. Predictive accuracy increases as meaningful insights drive better decision making, allowing them to adapt quicker to a future full of ambiguity. Although this future state has always been looming, the COVID-19 pandemic truly accelerated the global ability to adapt and react quickly to changing paradigm shifts. This new view of the world – and the technology and tools we now use – are now coming into focus.

Digital Twins, Geospatial AI Offer Deep Insight

Data is a challenge for infrastructure organizations across all sectors. Power, water, and communications must manage data on a growing scale, particularly as assets become more distributed and complex. This is where the idea of the “digital twin” comes to into play. A virtual representation that serves as the real-time digital counterpart of a physical object or process, a digital twin lives physically in both the real world and in the digital one.

Geospatial Artificial Intelligence (AI) is an emerging capability that will help fuel the use and understanding of digital twins. Loosely defined as using machine learning to understand the physical world at scale, the value of geospatial AI lies in its inherent ability to maintain the relationships between data and the real-world. When paired with connected data or graph technologies, this level of data can improve situational awareness in dynamic environments.

Let’s look at this capability from an aspiring smart city perspective on a simple but thorny example; the infamous pothole. In this context, the existence/location and general uncertainty of what is actually in the field and what will be required to fix it, tends to add a significant amount of time to the actual tasks to be performed. By relying on digital twins and machine learning, smart cities can automatically update pothole inventories and cross reference with traffic patterns to more effectively plan maintenance activities, drive down cost, and focus on repairs that provide the best value and quality life to the city. Even more exciting is that over time, patterns can be identified that may point to more underlying fundamental problems that could influence road construction techniques or environmental controls to reduce or eliminate the issue in the future.

For example, if a field planner needed to inspect in a certain geographic area, and had geospatial AI and digital twin capabilities, the team would simply press a button and in a couple of hours, have its answer – both to get the work done and insights to management for reporting. So, the idea is to go and collect this data in advance, primed and ready for use. This example is from a city perspective, but the same idea could be applied to any street side infrastructure.

In that situation, investing in passive data collection could be as simple as installing sensors and dashboard cameras on field operations vehicles. Vehicles driving down the road could capture the majority of what constitutes a city or region’s infrastructure. The field crew would drive its route, automatically collecting and recording data, with repeated data collection providing planners with a comprehensive understanding of temporal changes to infrastructure, deepening the quality of data.

This capability could even be applied to other use cases. For example, utilities using geospatial AI technology could send out their field ops trucks after a storm to quickly and easily record any damage to their infrastructure. The system’s deep learning ability would alert to any change by sending an automatic notification to the field crew via mobile device. Plus, this level of deep learning offers targeted data, flagging perhaps only 10 minutes out of 24 hours of data that is worth analyzing.

Cash-strapped organizations such as public agencies and utilities could benefit from this approach by offering their municipal fleets and field ops trucks up for service. By adding sensors and dashboard cameras, these organizations could monetize this data, opening up another potential revenue stream.

Example: BV Safe Contact App

Last year, Black & Veatch introduced its BV Safe Contact app, which relies on geospatial AI to serve as a tracking tool to help protect field service workers and construction crews during the COVID-19 pandemic.

The tool takes contact tracing to another level by combining real-time mapping technology with active COVID-19 case data which is streamed directly from Johns Hopkins University, and linking it to active work sites to continuously monitor potential exposure. Field crews can then make informed decisions on where to allocate resources, allowing them to relocate people away from high-risk areas and over to safer, less-affected areas.

For example, even though a construction manager may only have direct contact with a small number of people, that small group of people may expose the whole team. The app relies on Black & Veatch’s Nora, a cloud-native connected data platform that the company deployed last year to manage graph visualizations and analytics on large distributed infrastructure projects, to offer a three-dimensional view on how data, people and projects are interconnected.

The app is currently being used by a large northeastern utility that serves more than three million customers to make informed decisions on where, when, and how to schedule site work and allocate teams.

Closing the Digital Divide

These capabilities also hold incredible promise when it comes to closing the digital divide. In 2019, House bill 2643 hit the table requiring “the Federal Communications Commission to establish a challenge process for collecting and using broadband coverage data submitted by private and governmental entities to verify coverage data reported by broadband service providers.” In short, the bill decreed that telecom carriers had to provide data that illustrates how much of their infrastructure is available to those who are underserved.

Unfortunately, even today, this information is severely lacking, if not non-existent. There could be entire census blocks that are determined to have gigabit broadband because one small internet company rented a space, and therefore that entire census block is considered gigabit access.

Implementing digital twins and geospatial AI will answer several critical questions, namely, are these assets truly in the field, and where are they? How large is  the divide between where we are now, and we need to fully reach and serve those underserved people? Where can carriers start to bridge this divide? Information needs to be collected and prioritized in order to make these types of decisions and answer these questions.

With the Biden Administration’s recently announced infrastructure investment plan, The American Jobs Plan, there is new opportunity to see an infusion of federal money to address the digital divide. If passed, the plan would allocate $100 billion to expand broadband throughout the country.

According to the White House, the plan would lower the cost of internet, expand access in rural and urban areas, hold broadband providers more accountable and save taxpayers money. The plan would also require that providers disclose their monthly internet pricing to help remove barriers to competition and level the playing field for providers.

Turning this level of funding into success will require understanding what carriers currently have in the field, how they can create the greatest value, and how this data can relate to the end goal – all of which will require digital twins and geospatial AI.

Conclusion

In the past, none of this was possible for a variety of reasons, but primarily because the costs of the sensing technology, e.g., the cameras and monitoring devices, was simply too high. But economies of scale have driven down costs across the board; mobile phones in the commercial markets are now far more affordable, cloud technology has matured to the point where it’s become a fundamental cornerstone, and the cost of entry is pretty much pay as you go.

Digital twins and AI analysis would offer significant benefits to organizations across all sectors. By providing a comprehensive look at a geographical area and its infrastructure and assets, these technologies will enable smarter and more targeted field planning optimization. It could help digitize field surveys, offer new levels of remote engineering access, and enable contact tracing around COVID-19.

The focus will continue to shift away from the data itself and towards its relationships. The connections between data are where the most powerful insights lie. With enough data points, organizations can look to analytics to better understand the context and “see” the future.

AI at scale and emerging data technologies truly illustrate this connectivity and potential. Although it’s an emerging field, the benefits are limitless.

Related Articles

Responses

UA-171086932-1