Physix Frontier · News Briefing Card (Leiphone · Nov 14, 2026)

ZJU's Peng Sida Shifts from 3D Reconstruction to Embodied Navigation

KEY FACTS

  • After earning his PhD in 2023, Peng Sida stayed on at Zhejiang University as a researcher at the School of Software Technology.
  • His team publicly released two robot navigation works, SuperNav and ScaleNav, in October.
  • SuperNav handles decision-making and planning, while ScaleNav handles low-level locomotion control.
  • Street Gaussians has been applied to Li Auto's autonomous driving simulation system.
  • Peng Sida believes that in embodied intelligence, the action modality is more critical than the visual modality.

PHYSIX OBSERVATION

Peng Sida's choice of route is noteworthy: rather than chasing the end-to-end large-model hype, he uses a general-purpose VLM plus an Agent that plugs in cerebellum-style tools, sidestepping the communication-latency bottleneck of local large-model deployment. This is a reminder for embodied-intelligence startups: stacking parameters is not as good as clearly separating perception, memory, and action; engineering feasibility may determine the speed of deployment more than model scale.

Source: Leiphone report