by Jairam Ranganathan

How generative AI will benefit physical industries

feature
Feb 19, 20246 mins
Artificial IntelligenceGenerative AISoftware Development

Generative AI will reshape how we develop AI-driven products for the physical economy, starting with the creation of synthetic data sets for challenging use cases.

virtual eye / digital surveillance, privacy / artificial intelligence / machine learning

In recent years, artificial intelligence has undeniably revolutionized various sectors in the digital economy such as retail, customer service, and even art. Large language models, like ChatGPT, are changing communication and have offered innovative solutions for businesses. However, there is a critical segment of our economy that has yet to fully embrace AI’s potential—the physical economy.

The physical economy encompasses the industries that transport goods, power our homes, cultivate our food, and maintain the infrastructure that keeps society running. This includes sectors such as transportation and logistics, construction, energy, field service, and more. Just as generative AI tools have transformed consumer-focused applications, they also can substantially reshape how we develop AI-driven products for the physical economy.

One of the most pressing challenges confronting companies operating in the physical economy is safety. Within this context, generative AI applied to computer vision may emerge as one of the most significant advancements with the potential to reshape these industries for decades to come.

Building highly accurate AI models for computer vision

Building highly accurate AI models to detect a large variety of behaviors, particularly when it comes to physical workers, requires a lot of data. The challenge here is that the scenarios for which this data is needed are often dangerous and challenging to source. This is where the power of generative AI goes from useful to indispensable.

Unlike discriminative AI models that make predictions based on existing data, generative AI synthesizes entirely new data. These synthetic data sets can effectively train models that are difficult, if not impossible, to build through real-world data sources due to sparsity, complexity, or even danger. For instance:

  • To create a model for alerting drivers to traffic violations using traditional data sourcing, one would have to commit these violations, record them, and create a data set. This process is inherently dangerous, not to mention time-consuming and expensive. But according to the Federal Highway Administration, more than 50% of crashes with fatalities or injuries occur at or near intersections due to issues like these, highlighting the importance of finding solutions.
  • In a similar vein, developing AI models for predicting and detecting critical scenarios such as collisions requires sourcing data that captures such scenarios. Simulating dangerous conditions like a deer running into the road, an equipment failure on a construction site, or rocks sliding down a hillside toward a vehicle are very challenging to accurately replicate for the sake of creating training data sets.
  • Being able to accurately detect and flag when a high-value piece of machinery is malfunctioning or being mishandled can have a tremendous impact on worker safety. “Struck-by” deaths on job sites are a leading cause of fatalities on work sites and an estimated 75% are caused by heavy machinery. But building robust AI-powered detection to monitor this requires sourcing data for motion, mechanical operation, and worker usage, making it a complex and multifaceted challenge.

Generative AI allows us to generate realistic, synthetic data sets for diverse and challenging use cases. Developers can incorporate ancillary data and additional context such as road conditions, job site conditions, geo-location, customer service interactions, and other inputs to create rich data sets, with which we can train new models capable of detecting and alerting issues without the need for actual incidents to occur.

Unlock generative AI’s potential in physical businesses

To make generative AI’s potential a reality for a physical business, two crucial elements come into play: people and data.

Investing in a highly skilled team is a given precondition for success with any business. Also critical is having a diversity of expertise, as well as a diversity of experiences, cultural touch points, and background. Drawing on this expertise and experience to inform how generative AI is developed allows more context to be built-in, and the models can be expanded to serve a global audience versus a regional or national one.

Data quality in both edge computing and generative AI models is crucial. This is what has driven Motive to invest in a truly world-class annotations team. Because accuracy is so critical for the safety and optimization of our customers, this team ensures that the processes behind our use of generative AI are strong and consistent. These processes include ensuring the highest quality data and labels to train our models, and thus our products and services.

At the same time, generative AI in the physical economy will only be as useful as the insights and capabilities it creates. At Motive we use these insights and capabilities to power a comprehensive platform that provides insights on fleet and spend management, safety, asset monitoring, emissions, and more. This customer-facing technology element ensures that all of the work the teams put into their processes translates to something that drives meaningful results for a business and its clientele.

The transformative potential of generative AI

Generative AI has the potential to be transformative for the physical economy and the industries that fuel our everyday lives. What if generative AI could mitigate the impact of a wildfire by predicting its path and alerting residents sooner? Can predictive models around energy use be created to combat climate change and create more sustainable cities? Can shipping routes be improved by faster and more effective alerts of changes in weather or road conditions, or perhaps with fuel-efficient routes? These are questions that the use of generative AI can help us tackle for this key part of the global economy.

Jairam Ranganathan leads product management, design, data science, and strategy for Motive. Prior to joining Motive, Jai worked at Uber, where he served as senior director of product and data science, managing machine learning and AI, data, marketing systems, and operations tooling. Prior to joining Uber, Jai served as senior director of product management at Cloudera. He earned his BS and BA in computer science and mathematics at the University of Texas at Austin and completed his MS in computer science at Stanford University.

Generative AI Insights provides a venue for technology leaders—including vendors and other outside contributors—to explore and discuss the challenges and opportunities of generative artificial intelligence. The selection is wide-ranging, from technology deep dives to case studies to expert opinion, but also subjective, based on our judgment of which topics and treatments will best serve InfoWorld’s technically sophisticated audience. InfoWorld does not accept marketing collateral for publication and reserves the right to edit all contributed content. Contact doug_dineley@foundryco.com.