【RoboFi Market Report】Overview and case studies of emerging categories of robots and AI agents collaborating, managing, and economizing on the blockchain.
This could be the foundation for the coming Physical AI era after AI agents.
Good morning.
I am mitsui, a web3 researcher.
Today's "RoboFi Market Report" starts with an overview of RoboFi, then goes on to introduce specific projects, and ends with an overview of the challenges and prospects, so be sure to stay tuned for the end!
🤖What is RoboFi?
🗺️ specific use cases
💬 Seems like an essential feature in the future to come.
🧵TL;DR
RoboFiis a new category of robots and AI agents that are coordinated, managed, and economized on the blockchain, a concept similar to DePIN and DePAI.
The goal is to create a "robot economy" where robots have IDs, record their operating history in a chain, and perform tasks, make payments, and share data via smart contracts.
Issues to be resolved include data scarcity and rights processing, quality verification, interoperability, capital costs, security, and infrastructure dependence.
By leveraging blockchain to overcome these challenges, it has the potential to become the foundation for the Physical AI era that follows the age of AI agents.
🤖What is RoboFi?
As the name suggests, "RoboFi" is a coined word that combines "Robot" and "Finance" and is a new category for the coordination, management, and economization of robots on the blockchain.
I don't think it is that well known and established as a category yet. A similar category is DePIN or DePAI (DePIN+AI).
In this article, we will proceed by defining RoboFi as a project that crosses "Robot" and "Finance" and utilizes the blockchain.
◼️Market challenges and solutions
We will begin by describing the market challenges that RoboFi solves.
Traditionally, robots have operated in a closed environment for each manufacturer, with data and skills siloed. However, advances in AI and sensors, as well as the expansion of DePIN, are accelerating the movement to make robots function as "members of the network.
In RoboFi's view of the world,
From household cleaning robots to factory arms, delivery drones to humanoids,
Each has an ID, and the operating history is recorded in a chain,
Order, receive, and settle work via smart contracts with other robots and humans,
Learning data and skills can be shared and reused.
This "Machine Economy" will be realized.
To achieve this, we specifically aim to resolve the following issues
Lack of data and rights processing
There is still an absolute shortage of the behavioral and environmental data that is essential for robot learning. Furthermore, if the rights process is unclear, companies cannot use the data with confidence.
→ By blockchainRecord data contribution history and rightsand then solve the problem by automatically distributing licenses and rewards via smart contracts.Difficulties in quality verification
The performance of trained models is greatly degraded when spurious or low-quality data is introduced.
→ On-chain verification and evaluation protocols (Proof of Quality) allow only trusted data to circulate.Lack of interoperability
Currently, different manufacturers have different operating systems and standards, making it difficult for robots to cooperate with each other.
→Common OS and distributed coordination layerand DID-based robot ID standards enable interoperability of different aircraft.Cost of Capital and Monetization Barriers
The high initial investment required to introduce robots and the time it takes to recoup the investment are impediments to their widespread use.
→DAO the robot and distribute the operating revenue in tokens.or by converting robots to RWAs to allow for split investments, both raising capital and returning it to users can be achieved.Guarantee of trust and safety
Robot malfunctions and accidents pose a social risk.
→ Operation logs and diagnostic dataRecorded in the chain in a tamper-proof formand can be used for auditing and insurance to ensure transparency and trust.Concentration risk of infrastructure dependence
Relying on centralized services for infrastructure such as positioning (GPS), mapping, and communications can immediately result in usage restrictions and disruptions that ripple throughout.
→DePINcreates a distributed and user-owned infrastructure.
In other words, RoboFi is a new technology that will help us move from a world where "robots work by themselves" to a world where "robots are not a part of the system".Robots cooperate and participate in economic activitiesThis is a fundamental area for the creation of a "world". He then argues that in that future, the following challenges could exist, which is why blockchain will be utilized.
data deficiency
rights handling
interoperability
capital cost
safety
infrastructure-dependent
RoboFi solves these challenges using blockchain to realize the Physical AI era, which is next to AI agents.
🗺️ specific use cases
Now, let's look at some specific use cases to get an idea.
This article will be explained using the following chaos map as a guide.
◼️Learning and Training Infrastructure
This area is the foundation for supporting the learning of robots and AI agents.
For robots to operate in the real world, they require large amounts of data and a training environment. Learning and training infrastructure projects utilize human operating data and simulation environments to refine robot behavior through imitation and reinforcement learning. In this area, projects that allow people around the world to contribute to data collection and training through decentralized mechanisms and incentives fall into this category. This is the so-called "foundation" area for nurturing the intelligence of robots.
NRN Agents(Neuron)
It is a platform for training and competing AI agents in both the gaming and robotics domains. It uses human play data for imitation and reinforcement learning, and provides AI Arena, a system that pits AIs against each other in a game. The human players are rewarded with native tokens $NRN for the behavioral data they provide.
We are also working on continuous robot learning and plan to accelerate AGI research involving physicality through a Sim2Real learning pipeline and robotic sports competitions (Robotic Sports), where robots travel back and forth between virtual environments and real robots.
Keep reading with a 7-day free trial
Subscribe to web3 Research to keep reading this post and get 7 days of free access to the full post archives.