Config, a four-year-old startup based in Seoul, just closed a $100 million Series B round led by Korea’s biggest industrial names. SK Hynix, Samsung SDI, and LG Energy Solution—three of the world’s largest electronics and battery manufacturers—joined the round, along with existing backers like SoftBank Ventures and Hyundai Motor Group. The money will bankroll Config’s push to become the TSMC of robot data: the go-to supplier for the digital fuel that makes robots smarter and more reliable.

Instead of building robots themselves, Config sells the datasets and simulation environments that teach robots how to move, grasp, and navigate the real world. Think of it as the difference between Ford making cars and Bosch supplying the sensors and software that make those cars work. The startup’s software lets manufacturers plug in their own robot hardware and get it up to speed fast. That’s a big deal because most robot companies today waste months or years collecting and labeling their own data.

Why Korea’s manufacturers are betting on Config

SK Hynix, Samsung SDI, and LG Energy Solution don’t just make chips and batteries—they also assemble the robots that sort, pack, and inspect their own products. For Samsung SDI, that means robots placing battery cells inside electric vehicle packs. For SK Hynix, it’s machines handling ultra-thin semiconductor wafers. The problem they keep hitting is training those robots safely and quickly. Real-world robotics mistakes cost millions in damaged goods and downtime. Config’s data pipelines let them skip the trial-and-error phase by feeding robots simulated environments that mimic their factory floors.

The startup’s secret sauce is its library of synthetic data. It generates thousands of hours of robot movements in a virtual factory, including edge cases like a robotic arm colliding with a conveyor belt or a sensor failing mid-cycle. Those scenarios are nearly impossible to capture in the real world without risking expensive equipment. Config then labels every pixel and motion in these simulations, creating training datasets that can be dropped straight into a robot’s learning system.

How it compares to the AI chipmodel

TSMC doesn’t make chips itself—it makes the process of making chips so predictable and scalable that every major chip designer relies on it. Config is trying to do the same for robotics. Instead of selling robots, it’s selling the data pipeline that turns a generic robotic arm into a precision tool. Hyundai Motor Group, another investor, already uses Config’s software to train warehouse robots that sort and palletize car parts before they reach assembly lines.

Industry analysts say this model solves a bottleneck that’s stalled robotics adoption for years. Most robotics startups burn cash on custom hardware, but even the best hardware is useless without good training data. Config’s approach lets manufacturers focus on building better robots while outsourcing the messy work of making them actually work.

The road ahead

With the new cash, Config plans to expand beyond Korea into Japan and Europe, where industrial robotics markets are growing fast. It’s also doubling down on partnerships with robotics hardware makers, offering plug-and-play datasets for everything from collaborative arms to autonomous mobile robots. The startup claims its customers cut robot training time by up to 70%, which could shave months off product development cycles.

But the real test will be whether Config’s data can handle the messy unpredictability of real-world factories. So far, the startup has focused on controlled environments like warehouses and assembly lines. The next step is proving it can scale to more chaotic settings—think construction sites, hospitals, or even homes. If it pulls that off, Config might just become the invisible backbone of the robot economy.

What’s clear is that Korea’s industrial giants see something in Config that Silicon Valley’s robotics unicorns missed. They’re not betting on the next flashy humanoid robot. They’re betting on the boring, essential data that makes every robot work.

What You Need to Know

  • Source: TechCrunch
  • Published: May 11, 2026 at 10:58 UTC
  • Category: Startups
  • Topics: #techcrunch · #startups · #venture-capital · #korea · #config · #instead

Read the Full Story

This is a curated summary. For the complete article, original data, quotes and full analysis:

Read the full story on TechCrunch →

All reporting rights belong to the respective author(s) at TechCrunch. GlobalBR News summarizes publicly available content to help readers discover the most relevant global news.

Curated by GlobalBR News · May 11, 2026

🇧🇷 Resumo em Português

A Coreia do Sul acaba de dar um salto estratégico no futuro da robótica ao investir US$ 100 milhões em uma startup que promete ser para os dados de treinamento de robôs o que a TSMC é para os chips. A Config, que já é chamada de “TSMC dos dados robóticos”, acaba de garantir o apoio dos maiores conglomerados industriais sul-coreanos, sinalizando uma corrida global pelo controle da infraestrutura crítica que alimentará a próxima geração de máquinas inteligentes.

O Brasil, embora ainda engatinhando nesse ecossistema, não pode ignorar o movimento: o país precisa urgentemente entender como essa revolução impactará setores como manufatura, agricultura e logística — justamente aqueles em que nossa economia tem peso global. Afinal, se os dados são o novo petróleo, como já dizem os analistas, a Config está construindo os poços que vão abastecer as veias da indústria 5.0, e quem dominar essa cadeia ditará as regras do jogo pelos próximos anos.

A próxima etapa deve mostrar se outras regiões, incluindo o Brasil, vão acelerar para não ficar para trás nessa nova fronteira tecnológica.