Style3D Takes Embodied AI Push to WAIC 2026 With SynReal World and StyleClaw

The company is extending its clothing simulation stack into robot training data and light-industry AI workflows, but product availability and commercial terms remain undisclosed.

Style3D, formally known as Lingdi Technology, used the 2026 World Artificial Intelligence Conference in Shanghai to place its booth in the embodied intelligence exhibition area for the first time, presenting a simulation data engine for embodied AI and an AI workflow platform for light industry.

The company showed SynReal World, which it describes as a full-stack simulation data engine for embodied intelligence, alongside StyleClaw, an AI digital-partner platform aimed at apparel and other light-industry operations. A store robot was also displayed to demonstrate offline service interactions, although Style3D did not identify whether the robot was developed in house, supplied by a partner, or used solely as a demonstration unit.

The move marks a material product expansion for a company whose core business has been 3D garment design and clothing simulation since its 2015 founding. Style3D’s established software focuses on difficult flexible-material problems, including stretch, drape, folds and collisions. The company’s chief scientist, Wang Huamin, previously taught computer science at Ohio State University, and his team’s work on real-time GPU cloth-collision simulation was accepted at SIGGRAPH 2022.

At WAIC, SynReal World was presented as covering four stages of an embodied-AI data loop: 3D asset generation, spatial scene construction, virtual training and simulation-based evaluation. That positioning puts Style3D in the growing infrastructure layer around robotics, where developers are seeking ways to generate more varied training data without relying entirely on expensive physical data collection.

For global buyers of robotics software and industrial AI tools, the important question is whether Style3D can transfer its accuracy in deformable-object physics into broader robot environments. Apparel simulation is relevant to robotic tasks involving textiles, soft goods, packaging and other objects that do not behave like rigid industrial parts. If SynReal World can model those conditions reliably, it could be useful to manufacturers building automation for fashion production, retail fulfilment, warehousing and light assembly.

The transition is not assured. Simulated robot performance often weakens when models move into real settings, a problem commonly known as the simulation-to-reality gap. Style3D’s background in cloth physics may help in flexible-material scenarios, but the company has not provided public evidence from deployed robot training programs or comparative simulation benchmarks.

StyleClaw represents the more immediate application-layer expansion. Style3D said the platform supports design, product selection, marketing and operations in apparel and related light-industrial settings, with the ability to call tools and feed results back into workflows. The company has accumulated nearly 1,000 enterprise customers over roughly a decade and says it covers digital processes from design through production. It also cited an 80% customer renewal rate.

That installed base could matter commercially. Rather than selling AI tools into entirely new industrial accounts, Style3D may be able to layer workflow automation onto existing apparel digitization relationships. For supply-chain operators, the potential value lies in linking design decisions, product development, merchandising and production information within a common data environment.

Key commercial details remain unclear. Style3D has not publicly disclosed whether SynReal World and StyleClaw are formally available to external customers, their pricing models, rollout timetable or early deployment volumes. The company also has not disclosed the ownership or supplier structure behind the store robot shown at the booth. Those details will determine whether the WAIC presentation develops into a software business with production-scale customers or remains an early technology demonstration.