Markov Robotics Innovates Dexterous Manipulation with Video-Based Policy on LTX-2.5
- Published
- Source
- HumanoidsDaily
Markov Robotics has introduced a novel approach to dexterous manipulation using a video model as the policy for its LTX-2.5 robot. This method allows the robot to understand object interactions in real-time, enhancing its ability to perform tasks in unfamiliar environments. By leveraging a world model that incorporates physical knowledge, the robot can adapt its grip based on the object's characteristics, such as applying gentle pressure for fragile items like glass and firmer grip for sturdier objects like rocks.
This innovation addresses a significant challenge in robotics, where the development of effective software has not kept pace with advancements in hardware. Markov's approach emphasizes the importance of policies in robotic performance, suggesting that the future of dexterous manipulation lies in integrating high-fidelity sensory data with video-based expectations. The company's method of using a head cam to capture scenes and generate action videos represents a shift in how robots can learn and execute tasks.
Looking ahead, the robotics community will be watching how Markov's techniques influence the broader field, particularly as more organizations explore world models for robotic applications. The effectiveness of this approach and its potential to streamline the training process for robots could set new standards in the industry. No further timeline was disclosed at the time of publication.
Original report
The Policy Is a Video: How Markov Robotics Runs Dexterous Manipulation on LTX-2.5
HumanoidsDaily · [email protected] (Humanoids Daily Staff)This briefing is an independently written summary based on publicly available reporting and is provided for industry information and news discovery. The original report and source publication are credited and linked where applicable. RobotToday does not claim ownership of third-party source material.
Rights concerns? Contact [email protected] with the relevant URL and details. We will review the matter and take appropriate action where warranted.
More related news
-
1
RobotToday Weekly October 26 – 30, 2026 RobotToday · 1 day ago
- 2
- 3
- 4
- 5
- 6
- 7
- 8
- 9
- 10
- 11
- 12
Related Suppliers
Rhoda AI
United States
Palo Alto robotics company building a video-prediction robot control model (DVA, FutureVision) and a wheeled 25 kg-payload robot; USD 450M Series A in 2026.
Light Origins 亮源新创
China
Shenzhen physical-AI startup founded in late 2024 by ex-OpenAI researcher Roger Jiang, building the Light-O1 embodied foundation model and a bipedal humanoid.
Fixtureworks
United States
US manufacturer and supplier of modular workholding, fixturing components and standard machine elements for machining, inspection and automation.
Autonomique
Canada
Builds physical-AI software for industrial robots, enabling autonomous contact-rich assembly and dexterous manipulation on production floors.
ShengShu Technology 生数科技
China
Beijing-founded AI company building general world models; its Motubrain product is a world-action model for physical robot control and embodied intelligence.
Manifold AI 流形空间
China
Beijing embodied-AI startup that introduced its mission in May 2025; builds the RoboScape and WorldScape world models for robot control.
Runway AI, Inc.
United States
US AI research company building general world models (GWM-1) for video generation and a robot-learning platform for policy inference, sim evaluation and data.
Claru
United States
Reka AI's training-data unit, capturing egocentric video with 10,000+ wearable-camera collectors plus teleoperation and manipulation data for robot learning.
Beijing Simulation World Engine Co., Ltd. (SimWE) 北京云道智能科技股份有限公司
China
Beijing CAE developer (est. 2014) of the Simdroid multiphysics platform and Sim-PI, a GPU physics-simulation platform for training embodied robot models.
World Labs
United States
San Francisco spatial intelligence AI company founded 2024 by Fei-Fei Li; builds Large World Models that generate navigable 3D environments from text or images.
Elumotion Ltd
United Kingdom
UK robotics engineering company specializing in the design of dexterous robotic hands, wrists and humanoid torso systems that replicate biological motion.
Neurocean Technologies Inc. 深圳忆海原识科技有限公司
China
Shenzhen developer of brain-inspired AI and dexterous manipulation robots, including a modular 25-DOF industrial-grade dexterous robot hand.
Robotics needs a service framework.
RSF defines a common language for robot service capability, lifecycle operations, certification pathways, and service-provider networks.
Join the RobotToday community on LinkedIn
Daily robotics news, in-depth analysis, conference highlights, and discussions with professionals worldwide.
