Varun Vijaykumar Vupparige : Making Tyres Talk: The Push To Predict Grip Before It’s Lost

For most of the modern auto era, engineers have been able to describe a car’s behavior with impressive precision, from engine output to brake force to steering response. Yet one of the most important variables in safety has remained difficult for a vehicle’s computers to pin down: how much grip the tyres actually have on the road at any given moment.

That uncertainty lives in the contact patch, the palm-sized area where each tyre meets the pavement. It is small, constantly changing, and decisive, especially on wet highways, in snow, or on mixed surfaces. Varun Vijaykumar Vupparige, an algorithm development engineer at The Goodyear Tyre & Rubber Company’s Innovation Center in Akron, Ohio, has been working on software intended to make that invisible variable measurable in real time.

Goodyear’s broader push is to turn conventional tyres into sources of usable data for advanced driver-assistance systems and, eventually, autonomous and software-defined vehicles. Vupparige’s piece of that effort focuses on tyre-road friction estimation: using computation to infer how close a tyre is to its traction limit, without relying on expensive new sensors embedded in the tyre.

“The car already produces an enormous amount of data, wheel speeds, steering signals, inertial measurements,” Vupparige said. “The question I care about is whether we can infer the friction limit from signals the vehicle is already generating, before the driver or the autonomy stack ever asks the tyre for more than it can give.”

The distinction matters because many existing approaches effectively recognize low grip only after the vehicle is already flirting with instability. In technical terms, traditional methods often become reliable when a tyre is operating near 60 to 70 percent of its available grip, conditions more typical of emergency maneuvers than everyday driving. Vupparige’s framework aims to estimate friction earlier, around 30 to 40 percent grip utilization, the range where normal acceleration, braking, and cornering typically sit.

In practice, that means shifting from a system that detects trouble to one that can anticipate it. If software can estimate available traction before a vehicle reaches its limit, safety systems can adapt sooner: reducing speed, altering following distance, or modulating control strategies, rather than reacting after a skid has already started.

To build the estimator, Vupparige used nonlinear observers, a class of algorithms in control engineering that reconstruct hidden states from measurable signals. He then focused on making the approach practical for automotive deployment, where code must run predictably on cost-sensitive processors. Vupparige says he reduced the algorithm’s computational overhead by roughly 70 percent while maintaining accuracy and stability.

He also helped develop the simulation stack used to test the estimator across a wide range of scenarios. Those model-in-the-loop and software-in-the-loop environments allow engineers to explore edge cases: rapid surface changes, varying tyre conditions, abrupt maneuvers, that are difficult or unsafe to reproduce repeatedly on a test track. A vehicle test campaign was used to compare simulation outputs against real telematics data.

The work is now attracting wider attention. Goodyear’s tyre-intelligence efforts were recognized at the Tyre Technology Expo 2026, and the friction-estimation framework is being evaluated by automakers and Tier-1 suppliers assessing how to incorporate traction awareness into next-generation safety and automation systems.

An industry figure familiar with tyre intelligence said the approach stands out because it avoids a common stumbling block: cost. “The idea of a tyre that talks to the car has been around for years, but it usually comes with a bill of materials nobody wants to pay,” the person said. “What’s interesting here is that it leans on estimation and clever math rather than embedding costly sensors in every tyre. That’s the version of the idea that can actually reach production.”

For Vupparige, the goal is to give automated driving software something experienced drivers develop over time: intuition about the road.

“A good driver feels the road getting slippery and backs off before anything happens,” he said. “We’re trying to give that intuition to the software, so that an automated system has the same early warning. The tyre is the only part of the car touching the ground, so it should be the car’s most honest sensor.”

As automakers move toward vehicles that make more decisions in software, traction estimation—turning rubber-on-asphalt physics into reliable, real-time signals could become a foundational capability. And for companies trying to improve safety without adding substantial hardware cost, that may be the most compelling part of the promise.