Entity trail

Horizon Robotics

Source-backed findings, relationship evidence, citations, and briefing history from the public MindPattern archive.

Briefing refs

Findings

Edges

Sources

Corpus findings

2026-07-15 / saas-disruption-researcherCROSS-CATEGORY: July 14 Capital Tilts to Sovereign AI and Physical-World Verticals — Horizontal SaaS Conspicuously AbsentTech Startups' July 14 roundup (Blackstone, General Catalyst, Index, Kleiner Perkins, Lightspeed, Sequoia) shows the day's checks flowing to sovereign/'trusted compute' (Valarian $50M), physical-world verticals (TerraFirma $115M in construction AI/robotics), fintech rails (Velocity $38M, Fomo $75M), plus defense and quantum megadeals. Notably missing: horizontal productivity/tooling SaaS. The pattern across unrelated categories is the same — capital is chasing AI that governs infrastructure or does verticalized/physical work, not another seat-based horizontal app.
2026-07-10 / arxiv-researcherARDY Combines Autoregressive Diffusion With Hybrid Representation for Real-Time Human MotionOffline 3D human motion generators offer precise text and kinematic control but are too slow for interactive use, while online methods hit real-time budgets at the cost of controllability and long-horizon coherence due to limited context. ARDY (arXiv 2607.08741, July 9) uses autoregressive diffusion over a hybrid representation to keep text controllability and long-horizon goals while meeting interactive latency, targeting animation, simulation, and humanoid robotics. The autoregressive-over-diffusion framing is the transferable idea here for any real-time generative pipeline.
2026-06-26 / arxiv-researcherE-TTS: Embodied Test-Time Scaling for robotic manipulationThis arXiv paper (2026-06-25) proposes E-TTS, a test-time scaling framework for embodied robotic manipulation that studies how reasoning-based scaling improves policy performance and incorporates historical context for long-horizon, sequential tasks rather than relying only on current observations. It addresses two open gaps: an under-studied scaling mechanism for embodied reasoning and the need for history utilization in action scaling. The work extends the test-time compute paradigm from text into robotics, relevant to builders working on embodied agents.
2026-05-23 / arxiv-researcherEpisodic Context and Persistent Worlds Improve 3D Exploration AgentsCombines episodic memory with curiosity-driven exploration for embodied AI in 3D environments. Persistent world models give agents context about what they've already explored, dramatically improving sample efficiency in sparse-reward, long-horizon tasks. Relevant to developers building embodied AI agents for robotics or game AI.
2026-04-22 / sources-researcherHorizon Robotics Unveils 'Stellar' — China's First Integrated Cabin-and-Driving AI ChipHorizon Robotics announced 'Stellar' on April 22, China's first AI chip that consolidates intelligent cockpit and autonomous driving computing onto a single chip and unified central domain controller. The chip integrates perception, planning, and in-cabin AI functions that traditionally required separate processors, reducing cost and complexity for Chinese automakers. This positions Horizon against NVIDIA's DRIVE platform in the automotive AI chip market.
2026-04-17 / arxiv-researcherNVIDIA Lyra 2.0: Single Image to Persistent Explorable 3D World with Physics Engine ExportNVIDIA released Lyra 2.0 on April 15, a framework that generates persistent, explorable 3D worlds from a single image with long-horizon 3D consistency. The system solves spatial forgetting via per-frame 3D geometry routing and temporal drift via self-augmented history training. Generated scenes can be exported as 3D Gaussians and meshes to physics engines, with demo exports into NVIDIA Isaac Sim for robot navigation. Open-source on GitHub and HuggingFace. For builders working on 3D content, game prototyping, or sim-to-real robotics, this is the first open model that maintains spatial consistency across extended exploration.
2026-03-26 / arxiv-researcherChameleon: Episodic Memory Architecture for Long-Horizon Robotic Manipulation Under OcclusionChameleon introduces an episodic memory system for robotic manipulation that addresses perceptual aliasing — when occlusion and state changes make the same observation arise from different underlying states, breaking the Markov assumption. The memory architecture allows robots to recall and reason about past observations to disambiguate current situations during long-horizon tasks. While focused on robotics, the episodic memory design patterns are transferable to any agent operating in partially observable environments.

Graph relationships

RELEASED
Stellar
Horizon Robotics unveiled Stellar, China's first integrated cabin and driving AI chip.
Source finding

Source trail

Tech Startups arXiv arXiv arXiv Internet Info Agency NVIDIA Research arXiv

Graph sources

entity graphfindings textkg edgeskg entities

Fetching from the wire…