Nankang Tire × Profet AI: From the Era of Fuel to the Age of “Digital Tires” — The Road to Traditional Industry Transformation
After six decades in business, why must Nankang Tire, a true reflection of Taiwan’s manufacturing industry, undergo transformation?
Founded in 1959, it has witnessed Taiwan’s evolution from OEM to brand creation and accompanied the automotive industry from fuel-powered to electric vehicles. Its product line continues to expand—ranging from passenger car tires to off-road and winter tires—with more than 3,000 specifications accumulated.
However, in today’s data- and algorithm-driven age, traditional methods are failing: products are increasingly diversified, delivery times shorter, and demand more volatile. Any small error can snowball into massive costs.
“We produce in small batches with high variety,” said Fu-Chieh Chang, Deputy Manager of Product Development, “and that complexity forces us to respond faster. If we still rely on manual scheduling and paper records, we’ll be dragged down by inefficiency.”
The root problem lies in scattered information, unshareable experience, and inconsistent communication—each link operates in isolation, lacking a platform that enables full-chain collaboration.
Thus, Nankang Tire began to reevaluate its entire process—from R&D to production, from inventory management to order forecasting—step by step identifying where automation and datafication could be applied.
The New “Kung Fu” of Traditional Manufacturing: How Nankang Tire Uses Data to Push R&D Limits
“Tire manufacturing is a tough business,” Chang admitted. “We’re squeezed between global giants like Michelin and Continental on one end, and low-cost competitors on the other. We have to survive in the middle.”
Nankang exports to 189 countries, with 90% of its production sold overseas. Yet efficiency and cost remain a constant tug of war.
“The R&D cycle for a new tire pattern is long—it takes one to two years from concept to market,” said Chang. “It involves design, mold trials, verification, and corrections—each step takes time.”
To shorten both R&D and production, Nankang Tire partnered with Profet AI to introduce AI into key processes. From R&D to marketing, Profet AI is now applied in four major areas:
1. Contact Patch Prediction
Previously, engineers had to draw, mold, and run a contact experiment to check how the tire touches the road—a process that took several days and significant manpower. “Sometimes it took half a month to get results, and if the contact was uneven, we had to start over,” said Chang.
Using Profet AI’s AutoML, Nankang imported thousands of R&D data points—covering mold design, dimensions, and weight—to build a contact patch prediction model. Now, engineers simply input design parameters, and within seconds, the model predicts contact length and width while comparing against historical best samples.
The experimental cycle shortened from 12 days to about 3.5 days, dramatically cutting mold-testing costs and manpower.
2. Simulating Snow with AI — Reinventing Rubber Compound Design
Performance depends not only on tread pattern but also on rubber compounds. “Our product range is vast, and matching the right compound is key,” said Chang. For the European market, winter tires must balance low-temperature flexibility with wet-surface grip—two traits that often conflict.
“Sometimes, when low-temperature stability improves, wet grip drops,” he explained. Taiwan’s warm climate and lack of low-temperature test facilities make R&D even harder.
With Profet AI, Nankang transformed past test data into predictive models that simulate how different compounds behave in cold conditions. By inputting formula ratios and material properties, engineers can now predict glass transition temperature (Tg) and modulus variation, helping identify key performance factors early.
Though minor deviations remain, the results show great potential: tests that once had to wait for winter can now be simulated virtually—speeding up R&D and market response.
3. Turning the Master’s “Touch” into Models — Tread Length Optimization
“This step requires adjusting length per specification. It used to rely on craftsmen’s intuition,” said Chang. “With so many specifications, human error was unavoidable,” often leading to scrap and rework, increasing costs.
Through Profet AI, the system now automatically recommends optimal parameters based on environmental and machine conditions. When changing specifications, operators input batch and temperature data, and the AI provides recommended values and deviation alerts.
Analysis showed that inaccurate settings once caused annual rework losses of over one million NTD. With AI, Nankang expects significant waste reduction and to turn personal know-how into collective factory intelligence.
4. Forecasting Orders with AI — Smarter Scheduling
Nankang also fed historical ERP order data into Profet AI for automated trend modeling. Each month, sales teams input recent data, and the system forecasts next month’s demand—supporting production planning and material preparation.
Early tests show the model can predict demand changes with promising accuracy. In the future, Nankang aims to anticipate hot-selling sizes, streamline production flow, and optimize resource allocation—reducing both idle time and waste.
“Traditional industries are like martial arts heroes,” said Chang with a smile. “They must constantly refine their inner strength through new techniques. For us, AI is that new ‘kung fu’—rebuilding our foundation.”
From Resistance to Co-Creation: How “Touch” and “Data” Learn to Communicate
For many employees, AI first appeared as an unfamiliar rival, fast and precise but speaking in data, not intuition. Craftsmen spoke the language of feel and experience, so friction was inevitable.
Chang admitted the hardest part was converting experience into data. In processes like calendering, factors such as room temperature, water temperature, and roller speed all affect results. “Some craftsmen say this compound needs to run slower, others say faster, but those terms ‘slow’ or ‘fast’ were never quantified.”
To address this, Nankang focused on mindset, mentorship, and motivation. R&D and line supervisors personally accompanied workers to fine-tune models. “That companionship matters, when they see results, the resistance fades,” said Chang.
But more important, he added, was a sense of achievement. When workers saw their expertise transformed into models and shared across the factory, they felt proud, their experience was no longer invisible but institutionalized.
To sustain this learning culture, Nankang established a skills assessment system, tracking employees’ learning and application progress. “Supervisors can see their team’s growth, and employees can see their own progress.”
“AI can be a tool, an assistant, even a consultant,” concluded Chang. For Nankang Tire, the goal of AI adoption is not just technical integration but the moment when people truly learn to move forward together with AI.