OpenRoboto Launched Decentralized Robotics Training Platform
The platform utilizes a tokenized incentive model to source egocentric video for vision-language-action robotics.
Updated on Sept. 22, 2026 in Robotics
Live Poll
Do you believe decentralized platforms are a reliable way to collect data for robotics AI?
OpenRoboto has launched Shift, a decentralized platform that compensates contributors for recording egocentric video to generate training data for robotics models. The company also introduced a native alpha token for its Bittensor Subnet 80, which facilitates distributed competition for robotics benchmarks.
Why it matters
The platform aims to solve the scarcity of structured, real-world training data for vision-language-action robotics models. By decentralizing the data collection process, OpenRoboto seeks to accelerate the development of models that rely on diverse physical environment navigation.
Subnet 80 miners compete against LIBERO-Pro benchmarks, with emission rewards structured in a 70/20/10 split. Axis Robotics has already contributed over 3 million multimodal trajectories to the project's Open Data Pool.
The players
OpenRoboto
An organization focused on decentralized AI and robotics development that operates Bittensor Subnet 80.
Axis Robotics
A robotics firm specializing in the development of multimodal navigation systems and training datasets.
Bittensor
A decentralized machine learning network that uses incentive-based subnets to coordinate AI training models.
Base
A secure, low-cost Ethereum layer-two blockchain network developed by Coinbase used here for token liquidity.
The details
Shift captures egocentric video footage—first-person perspective recordings of human movement through environments—and processes it into training data for vision-language-action models. These models map visual inputs to physical robotic control outputs. The ecosystem operates on Bittensor Subnet 80, utilizing ForeverMoney and Chainlink CCIP—a cross-chain communication protocol—to bridge the token to the Base blockchain.
Timeline
July 29, 2026: Subnet 80 went live on the Bittensor mainnet.
August 17, 2026: Axis Robotics contributed data to the Open Data Pool.
Early September 2026: 85% of emissions flowed toward physical-robot competitions.
September 17, 2026: Subnet 80 launched its native alpha token on Base.
September 22, 2026: The token holder count reached 1,909.
The Tech Race
OpenRoboto is positioning its decentralized data pipeline to compete directly with centralized laboratories using the LIBERO-Pro benchmark to validate model progress. The project marks a shift from monolithic datasets toward community-driven, incentivized video collection for physical AI models.
Developers and contributors can currently interact with the Subnet 80 token, which trades at approximately $8.20 on Base. The platform requires participants to provide egocentric video data to influence the training progression of physical-robot models in future iterations.
The takeaway
The success of this platform hinges on whether decentralized, crowdsourced video data can match the quality of highly curated professional datasets. Observers should track the competitive performance of Subnet 80 models against established LIBERO-Pro baselines to determine the viability of this model.
Further reading
For more on how new architectures are changing the field, explore the latest trends in Robotics.
Source note: This article includes information reported by Crypto Briefing.
Live Poll
Do you believe decentralized platforms are a reliable way to collect data for robotics AI?
