By harnessing AI, real-world data, and driver behaviour insights, drivebuddyAI is reducing accidents today while building the future of intelligent mobility.
On the busy highways where fleets move the country forward, safety often depends on more than just the machine—it depends on the driver behind the wheel. And this is exactly where Ahmedabad-based drivebuddyAI is quietly making a difference.
What began as an effort to bring intelligence into driver behaviour is now shaping into something much larger. Today, the company’s AI-powered Driver Monitoring Systems (DMS) are already active in around 2,000 vehicles, working in real-world conditions across India’s logistics network. And this is just the beginning.
With an order pipeline of another 5,000 installations and a clear target of reaching 8,000 fitments by 2026, the company is preparing for its next phase of growth. Speaking to this publication, Mr. Nisarg Pandya, CEO of drivebuddyAI, said, the focus is not just on scaling numbers, but on building a stronger, more reliable ecosystem around fleet safety. To support this journey, the company is also rethinking how it builds its hardware. Moving away from in-house assembly, it is partnering with Chennai-based e-con Systems for contract manufacturing—while continuing to retain full ownership of its core technology.
It’s a shift that reflects a larger ambition: to scale faster, without losing control of innovation. Because in a space where safety, data, and intelligence come together, every detail matters.
Beyond Safety Today
What makes drivebuddyAI’s journey interesting is that it isn’t just solving today’s problems—it is quietly building for what comes next. While the immediate goal is to improve driver safety, the larger vision goes much further. The company sees its technology as a stepping stone towards advanced driver-assistance systems and, eventually, autonomous mobility in India. By combining AI with real-world driving data and behavioural insights, it is not only reducing accidents today but also preparing the ecosystem for a smarter, more connected future, he explained.
At the core of this effort is its DMS; but this isn’t a basic alert system. It observes, learns, and understands. Instead of simply checking if a driver’s eyes are closing, it looks at a range of behaviours—yawning, unusual head or eye movement, even subtle signs like scratching or loss of steering control.
Fatigue, for instance, is not treated as a single moment. It is tracked in stages. The system begins to raise alerts at an early level—well before the driver actually falls asleep—giving fleet operators and drivers enough time to step in and prevent a potential incident. It also keeps an eye on everyday risks—phone usage, seatbelt compliance, and where the driver’s attention is focused. Yet, what stands out is how these alerts are designed. They are not intrusive or overwhelming. Instead, they are meant to guide, not irritate—making it easier for drivers to accept and act on them.
This level of intelligence has been built over time. With over four billion kilometres of Indian road data and insights from more than 5,000 drivers, the system has learned from real conditions—not simulations. And that is perhaps its biggest strength. It understands India’s roads, its drivers, and its realities. And in doing so, it is helping shape a transport ecosystem that is not just safer—but also smarter and ready for the future, he mentioned.
Where It Began
The idea behind drivebuddyAI did not come from a lab. It came from the road. India’s highways carry millions of commercial vehicles every day. Trucks, buses, and logistics fleets keep the economy moving. But behind this constant motion lies a serious challenge—road safety. Long hours behind the wheel, driver fatigue, distractions, and unpredictable traffic often turn these journeys risky. This is where drivebuddyAI found its purpose.
Built on AI and shaped by real-world data, the system was designed to step into this gap—to support drivers, not replace them. By closely understanding how drivers behave on Indian roads, it brings together monitoring, analytics, and connected vehicle technologies into one platform.
Over time, this approach is doing more than just reducing risks. It is slowly encouraging a more disciplined driving culture—one where awareness improves, reactions become sharper, and safety becomes a shared responsibility.
From Idea to Impact
The journey of drivebuddyAI began with a simple belief—that technology could make India’s roads safer. Mr. Pandya, an electronics engineer, started the company in 2016, with serious development taking shape by 2018. From the very beginning, the focus was clear – to use AI and automotive technology to reduce accidents in commercial vehicles.
The inspiration came from global advancements. Companies like Mobileye, Tesla, Google, and Delphi had already shown how cameras and sensors could detect hazards and assist drivers. The idea was to bring similar capabilities to Indian roads. The first step was building a system that could look ahead—detecting obstacles and warning drivers of potential collisions. A basic version was developed and tested. But that’s when reality stepped in.
The system, trained on global datasets, struggled in India. It couldn’t reliably identify autorickshaws. It got confused by pedestrians in sarees. Faded lane markings, construction zones, and chaotic traffic patterns made it even harder. That was a turning point. Instead of trying to adapt global models, the team decided to start fresh—building AI trained entirely on Indian roads, Indian drivers, and Indian conditions.
As the system evolved, so did its purpose. It moved beyond just watching the road to understanding the driver. Features like fatigue detection and distraction alerts were added, eventually leading to a dual-camera setup—one focused on the road, the other on the person behind the wheel. What started as a simple idea has now grown into a system shaped by real-world challenges—and built for them.
Learning on Roads
Early pilots with cab operators in Pune, Adobe’s employee transport in Noida, and UPSRTC buses gave the team a clear view of real-world challenges. Each deployment helped refine the system and build meaningful data. A UPSRTC tender just before COVID-19 offered another opportunity, though it didn’t move forward due to the pandemic. Soon after, in 2019, funding from Roadzen—now its parent—helped accelerate development and scale the platform further.
Precision Through Learning
For drivebuddyAI, accuracy isn’t a one-time achievement—it’s a continuous process. Every alert is reviewed by a dedicated team, studying real-world data to understand what worked, what didn’t, and what was missed. Even unusual situations—like drivers trying to outsmart the system—become learning moments. This constant feedback loop has helped improve drowsiness detection accuracy from around 70% in 2021 to 92% in 2025, he mentioned.
At the same time, the company is closely aligned with evolving regulations. Working with Indian bodies like ARAI and validated internationally, the system has already achieved 92% accuracy against European standards—well above the benchmark for top safety ratings.
Real-World Impact
For fleet operators, the value of drivebuddyAI goes beyond compliance—it shows up in everyday operations. With live video from both the road and the cabin, managers gain real-time visibility, helping them respond quickly to risks like fatigue or phone usage. Just as importantly, the presence of the system itself encourages safer driving behaviour.
The results are clear. In some deployments, fleets have reported up to 80–85% reduction in drowsy driving incidents within six months. The system also provides video evidence during accidents or disputes, offering protection against false claims. In the end, it’s not just about monitoring—it’s about building safer roads, one driver at a time, Mr. Pandya signed off.