This interview analysis is sponsored by Motive and was written, edited, and published in alignment with our Emerj sponsored content guidelines. Learn more about our thought leadership and content creation services on our Emerj Media Services page.

In industries where seconds matter and margins are thin, delayed feedback, manual processes, and limited visibility make it incredibly hard to scale safety and efficiency. According to the American Transportation Research Institute (ATRI) 2025 operational costs report, the industry’s average cost of operating a truck hit $2.260 per mile in 2024, with non-fuel marginal costs rising 3.6% to a record $1.779 per mile, squeezing profitability amid high accident risks and downtime.

Truck fleet accidents represent a massive scale problem across regions. According to Federal Motor Carrier Safety Administration (FMCSA) data cited in truck safety analyses, there were 153,342 commercial truck and bus crashes in the U.S in 2025, underscoring the sheer volume in fleet operations. Fatalities highlight the human toll. FMCSA also notes that 3,781 large trucks and buses were involved in fatal crashes in 2025.  

Peer-reviewed research on driver monitoring systems underscores the importance of early risk detection in preventing accidents. A comprehensive review published in IEEE Transactions on Vehicular Technology finds that human error, including drowsiness, distraction, and panic, drives 90% of road accidents, directly exacerbating fleet challenges like delayed feedback and lack of real-time visibility. 

A peer-reviewed study from Wright State University citing the IEEE figure also shows that their edge AI system achieves 98.6% accuracy in detecting these risks using low-power in-cab hardware and operates reliably even in poor lighting, on bumpy roads, or with partial face views. 

Tested across diverse drivers and conditions, it proves scalable for large fleets facing ATRI’s record $2.260 per-mile costs and FMCSA’s 153,342 annual truck crashes, turning reactive reviews into proactive safety.

In a recent episode of the ‘AI in Business’ podcast, Emerj Editorial Director Matthew DeMello was joined by Hemant Banavar, Chief Product Officer at Motive.  In their conversation, Hemant discusses how edge AI can transform the physical economy by delivering real-time, safety-critical insights that prevent accidents, modernize operations, and make high-risk industries safer and more efficient.

Their conversation highlights two critical insights for the physical economy:

• Modernizing the physical economy with AI: Deploying AI-powered operations platforms to replace manual processes, improve fleet safety, and increase productivity across sectors.
• Driving safer operations with real-time AI: Leveraging edge AI to deliver instant risk detection and feedback to drivers to prevent accidents, improve safety, and unlock substantial cost savings.

Listen to the full episode below:

Guest: Hemant Banavar, Chief Product Officer, Motive

Expertise: Product, Artificial Intelligence, Data Science

Brief Recognition: Hemant leads product, design, data science, partnerships, and strategy at Motive, driving innovation across fleet management, driver safety, spend management, and workforce operations. In his past stints, he has worked with Uber, Stripe and Microsoft. He holds a master’s degree in business administration from the University of California, Berkeley.

Modernizing the Physical Economy with Edge AI

Hemant begins by explaining that in many physical operations, feedback is received hours, days, or even weeks after an event, making it largely ineffective. In environments where operators make split-second decisions, such as braking a vehicle to avoid a collision, timely feedback is critical, as delays can directly affect safety. 

In such situations, humans cannot provide feedback fast enough, and delays are unacceptable. He emphasizes that the disconnect between humans and feedback is why AI models deployed at the edge are essential: they can detect risks and alert operators in real time, allowing them to de-escalate dangerous situations immediately.

While these models are trained using human judgment, they operate autonomously to provide instant feedback. Hemant stresses that in these high-stakes environments, accuracy, reliability, and trust are mandatory, and that this principle guides how they build their AI systems.

He highlights that this is where edge AI becomes valuable, as it can deliver real-time feedback, enabling operators to react immediately. By providing insights when needed, edge AI can help prevent accidents and ensure workers return home safely, underscoring the importance of immediate, actionable data in the physical economy.

In another episode of the ‘AI in Business’ podcast, Adam Burns, Vice President of Network and Edge and Director of Edge AI Development Tools at Intel, emphasized that the real impact of AI emerges when insights are delivered at the moment decisions must be made—transforming data from a retrospective tool into a real-time operational capability.

“AI becomes far more valuable when data is processed in real time. When insights are generated as events occur, organizations can act immediately — improving outcomes, increasing efficiency, and preventing issues before they escalate. That immediacy transforms AI from a retrospective analysis tool into a core operational capability that enables entirely new ways of running physical systems.”

– Adam Burns, VP of Network and Edge, Director of Edge AI Development Tools at Intel

Listen to the full episode with Adam featuring these insights and more below:

For the sake of listener understanding, Hemant defines the “physical” economy – a relatively new term, relative to what is described as physical AI – as the part of the economy involving people who deliver groceries, stock stores, build infrastructure, and manage public transportation, essentially, sectors like trucking, construction, oil and gas, and public services, which together make up about 50% of GDP. 

Despite its size, investment in technology for this sector has been disproportionately under-funded, he says, estimating that around 30% of investment over the past decade is going toward solutions for the physical economy. 

As a result, many companies still rely on manual processes, pen and paper, and phone calls to run operations. Hemant highlights that his company is among the first to introduce technology into this space, building an AI-powered operations platform that helps physical-economy operators run fleets more safely, productively, and profitably.

Driving Safer Operations with Real-Time AI

Hemant further explains that edge AI works by placing devices directly in the customer’s environment, often in the cab or vehicle, to monitor both the road and the driver in real time and detect risks. These systems use video and telematics data, such as engine RPM, speed, and external temperature, which are critical for assessing risk. In a separate conversation on the ‘AI in Business’ podcast, Naveen Kumar — then Director of Financial Crimes at Walmart, and now Head of Insider Risk, Analytics, and Detection at TD Bank — highlights that the true value of video intelligence emerges when it is integrated with other operational data, creating a richer and more actionable view of risk in real time.

“The organizations doing this well use video to augment their other data, not replace it. When visual data is integrated with transaction logs, access controls, and operational systems, decisions become far more context-driven and the full picture becomes much richer. That integrated view allows teams to move beyond isolated signals and understand risk, behavior, and operations in a more complete and actionable way.”

– Naveen Kumar, Head of Insider Risk, Analytics, and Detection, TD Bank

You can listen to the full episode with Naveen Kumar featuring these insights below:

Building hardware capable of handling this across multiple vehicles and scenarios is challenging. He highlights Motive’s new AI Dash Cam Plus, powered by a Qualcomm AI processor that can run up to 30 AI models simultaneously, enabling real-time monitoring of numerous behaviors. 

The device also features hands-free communication, enterprise-grade reliability, and dual forward-facing lenses that allow accurate distance perception for improved collision detection. Hemant emphasizes that this advancement not only enhances current risk detection but also enables the monitoring of more complex behaviors, opening the door to the next stage of edge-based AI innovation.

Hemant highlights that customers in the physical economy prioritize reducing accidents and running safer operations, and edge-based AI enables this by providing real-time video monitoring, alerts, and actionable feedback to drivers, rather than delayed feedback. 

Similarly, in another ‘AI in Business’ podcast episode, Joe Troy, Senior Manager of Site Risk at Amazon, explains that as real-time video analytics mature, they evolve from simple monitoring tools into cross-functional intelligence systems that help organizations operate more effectively across the enterprise.

“AI-powered video is not about replacing human insight—it removes manual work so teams can focus on what actually moves the business. By automating detection and surfacing meaningful signals, these systems allow organizations to respond faster and make more informed decisions across operations. When implemented correctly, video intelligence becomes a cross-functional layer that improves training, safety, and overall operational performance.”

– Joe Troy, Senior Manager of Site Risk, Amazon

Listen to the full episode with Joe featuring these insights and more below:

Hemant also notes that independent studies show Motive’s AI alerts detect unsafe behavior 2–4 times more effectively than competitors’, helping prevent accidents and change driver behavior. Across customers, Hemant asserts that Motive’s technology has prevented roughly 170,000 accidents and is estimated to have saved 1,500 lives, with collision reductions up to 80%:

He goes on to cite an illustrative client story with robust and measurable results:  

“One of our customers, Ernest Concrete — an 80-year-old concrete company — came to Motive after their previous dashcam failed to capture two major accidents. After deploying our solution, they saw cell phone usage among drivers drop by 97% and distracted driving events fall by 83% within 13 months.

Overall, those improvements translated into $6.5 million in savings over that period, which is significant for a company operating in a highly margin-constrained industry.”

– Hemant Banavar, Chief Product Officer at Motive

Hemant goes on to share that another customer, a national home services company called South Wind, achieved $2.5 million in savings, including $2 million from reduced insurance costs and $500,000 from fuel savings, thanks to Motive’s real-time insights, fraud prevention, and operational optimization.