1. DePIN+AI builds a robot paradigm in the AI era
On February 27, Messari hosted a podcast on Building Decentralized Physical AI and invited Michael Cho, co-founder of FrodoBot Lab. In this podcast, Michael Cho focused on the opportunities and challenges of DePIN+AI in the field of robotics.
Soon after the fermentation of Messari, the concept of DePIN robot quickly became popular, and many discussions about DePIN robots began to flood.
Our industry observations this week will also focus on analyzing and discussing our analysis and observations on this track with everyone.
Before we start our discussion, let’s take a look at the development of artificial intelligence itself:
In computing power, Nvidia’s quarterly revenue has grown fivefold over the past three years;
In the bandwidth area, North American data center construction has also increased fivefold in the past three years;
In the energy sector, OKLO alone will need 12.0 GW and TerraPower will need 4.0 GW;
In the data field, large companies invest more than $500 million each year to purchase wholesale data for training AI models.
Under the premise of the overall global economic downturn, AI, as the main scientific and technological revolutionary technology in the next ten or even twenty years, is leading all players in this field (computing power, energy, data) to accelerate at an accelerated growth rate of several times each year.
While AI is developing at such a high speed, concerns about it are also increasing day by day. The reason is that if AI computing power (similar to the engine of a car), AI big models (similar to controllers and processors), AI energy (similar to oil and fuel), and AI data (similar to raw materials) are controlled by a few centralized large companies, then the future technological era may be completely controlled by a few large companies, which may lead to absolute centralization and centralization. By then, we may really open the biggest Pandoras box ourselves.
It is precisely because of the concern about such a centralized situation that a new track and direction is being hotly discussed, that is, DePIN+AI. We at DePIN ONE are willing to define it as DePAI, that is, DePAI=DePIN+AI.
How will DePAI help AI become more decentralized?
We will expand and analyze the main content of Messari’s podcast with Michael last month.
There are many pain points in current AI. Although they have a variety of functions, they are all processing superficial information such as text. Such information is cold and has no temperature, and lacks deep perception and understanding.
The DePIN network can serve as the “five senses” and “limbs” of AI.
The five senses help AI perceive the real world in an all-round way. Some developers are now using ioID and W3bstream to connect real-world devices to the blockchain and use zero-knowledge proofs to verify their real activities.
The limbs can help AI make accurate judgments based on its own perception and put them into action, thus realizing the training -> modeling -> automation system very well.
1. DePIN makes AI data more real and diverse
Unlike those large online AI models that are trained with large amounts of Internet data, DePIN devices can help AI interact with the real world and obtain more real and real-time data. Only data trained with this kind of data can enable AI+robots and other devices to develop true embodied intelligence.
Since DePIN is still in its early stages of development, there is currently no such large-scale infrastructure in the world, and there is no consensus on how to collect this data.
We believe that the data that DePIN+AI will collect in the future can be classified into the following three categories:
The first type is human operation data, which is the data generated when humans manually control the robot. This type of data is of high quality and can capture video streams and action labels - that is, what humans see and how they react accordingly. This is the most effective way to train AI to imitate human behavior, but the disadvantage is that it is costly and labor-intensive.
The second type is synthetic data (simulated data), which is useful for training robots to move in complex terrain, such as training robots to walk on rugged ground, and is very useful for some specialized fields. But for some tasks with a variety of changes, such as cooking, the simulated environment is not very good. We can imagine the situation of training a robot to fry eggs: the type of pan, the oil temperature, and slight changes in room conditions will affect the results, and it is difficult for the virtual environment to cover all scenarios.
The third category is video learning, which is to let the AI model learn by observing videos of the real world. Although this approach has potential, it lacks the real physical direct interactive feedback required for intelligence.
If these data are collected and supported, the embodied intelligent service capabilities of AI will definitely be greatly enhanced.
2. DePIN maximizes the capital efficiency of AI and is more conducive to the decentralization of AI from the source, rather than being a puppet of certain capital.
Unlike traditional AI models that rely solely on computing power, the implementation of intelligent robotics technology requires the deployment of physical devices in the real world, which poses a huge capital challenge.
Robots are expensive to build, and only the wealthiest companies can afford large-scale experiments. Even the most efficient humanoid robots currently cost tens of thousands of dollars, making them unrealistic for mass adoption.
General robotics AI is still a long way from mass adoption, given the challenges of hardware, data, and evaluation.
However, the introduction of DePIN technology has given everyone hope.
Because the scale and coordination of a decentralized network can effectively disperse the capital burden, it can help some small entrepreneurial teams to develop this technology. In order for general robots to be more efficient as soon as possible and closer to real people, the development of robotics technology should be decentralized rather than controlled by a few large companies. Instead of relying on a large company to pay for the construction of thousands of robots, it is better to put individuals who can contribute into a shared network.
Furthermore, DePIN accelerates data collection and evaluation.
Instead of waiting for a single company to deploy a limited number of robots to collect data, decentralized networks can run in parallel and collect data at a much larger scale.
For example, in a recent AI vs. human robot competition in Abu Dhabi, researchers from institutions such as DeepMind and UT Austin tested their AI models against human players. While humans still prevailed, the researchers were excited about the unique dataset collected from real-world robot interactions. This underscores the need for subnetworks that connect the various components of robotics. Even if full autonomy remains a long-term goal, DePIN technology has demonstrated tangible value from data collection and training to real-world deployment and validation.
On the other hand, the DePIN network is helping the implementation of AI robots with higher efficiency and lower costs.
A specific example is that FrodoBot Lab has worked with the DePIN project to secure two boxes of NVIDIA H100 GPUs — each box contains the computing power of eight H100 chips, which provides researchers with the necessary computing power to process and optimize AI models for real-world data collected from robot deployments. Without such computing resources, even the most valuable data sets cannot be fully utilized. It can be seen that through access to DePINs decentralized computing infrastructure, the Robotics Network can allow researchers around the world to train and evaluate models without being restricted by capital-intensive GPU ownership. If DePIN can successfully crowdsource data and hardware advances, the future of robotics may come sooner than expected.
3. DePIN is helping AI and AI embodied intelligence achieve more efficient business efficiency
AI agents like Sam (a traveling influencer bot with meme tokens) demonstrate new revenue models for decentralized robotics networks.
Sam runs autonomously, live streaming 24/7 in multiple cities, while its meme token increases in value.
This model demonstrates how intelligent robots powered by DePIN can sustain their own finances through decentralized ownership and token incentives. In the future, these AI agents can even use tokens to pay human operators for assistance, rent additional robot assets, or bid for real-world tasks, forming an economic cycle that benefits both AI development and DePIN participants.
expect
The development of embodied AI depends not only on algorithms, but also involves hardware upgrades, data accumulation, financial support and human participation.
In the past, the development of the robotics industry was limited by high costs and the dominance of large companies, which hindered the speed of innovation. The establishment of the DePIN Robot Network means that with the power of decentralized networks, robot data collection, computing resources and capital investment can be coordinated globally, which not only accelerates AI training and hardware optimization, but also lowers the development threshold and allows more researchers, entrepreneurs and individual users to participate.
We also hope that the robotics industry will no longer rely on a few technology giants, but will be driven by the global community to move towards a truly open and sustainable technology ecosystem.
2. DePIN Track Data and Observations
1. DePIN’s overall share accounts for only 0.1% of the trillion-dollar AI market
The number of DePIN projects has increased from 100 in 2022 to 1,170 in 2024, the market value has soared from US$5 billion to US$50 billion, and the active node rate has increased from 2% to more than 50%. However, DePINs overall share only accounts for 0.1% of the trillion-dollar AI market. It is no exaggeration to say that this track has a growth potential of 100-1,000 times.
2. DePIN’s financing amount increased, but the number of financings decreased
According to Messari data, DePIN financing growth was flat year-on-year, with a higher amount of financing in the first quarter of 2025 but a lower number of financings.
Q1 2024: $156 million in 62 rounds of financing.
Q1 2025: $159 million in 36 rounds of financing.
The data shows that there are fewer emerging early-stage startup projects, but mature DePIN projects are expanding in scale.
At present, the global share of leading projects in the DePIN field is still very small, and they are in the absolute early stages of the track.
The market share in the field of wireless transmission is 0.002% (leading project Helium), the market share in the field of computing is 0.03% (leading project Filecoin), the market share in the field of energy is 0.001% (leading project Daylight), and the market share in the field of identity authentication is 0.2% (leading projects Worldcoin and Anymal).
The agent-based artificial intelligence market in the AI track is expected to grow significantly in the next decade, from US$520 million in 2024 to US$196.6 billion in 2034, with a compound annual growth rate of 43.8%.
3. Grayscale releases Q2 quarterly report, focusing on RWA, DePIN and IP tokenization
Grayscale released its 2025Q2 quarterly report this week, focusing on RWA, DePIN, and IP tokenization. As a result, three new tokens were added to the Top 20, namely IP, SYRUP, and GEOD, while Akash Network, Arweave, and Jupiter were removed.
The report shows that this quarter, Grayscale will focus on tokens that reflect the non-speculative applications of blockchain technology in the real world. These tokens are divided into the following three categories: RWA (real world assets), DePIN (decentralized physical infrastructure) and IP (intellectual property tokenization).
Among the three assets Maple (SYRUP), Geodnet (GEOD), and Story (IP) added to the list of Top 20 assets in the second quarter of 2025, two are DePIN projects.
Geodnet (GEOD): Geodnet is a DePIN project that collects real-time positioning data. As the worlds largest real-time kinematic positioning (RTK) provider, Geodnet provides geospatial data with an accuracy of up to 1 cm, providing affordable solutions for users such as farmers. In the future, Geodnet may provide value to autonomous vehicles and robots. The network has expanded to more than 14,000 devices in 130 countries, and annualized network fee revenue has grown to more than $3 million in the past 30 days (an increase of approximately 500% year-on-year). It is worth noting that compared to other assets in the top 20, GEOD has a lower market capitalization and is listed on fewer exchanges, so it can be considered riskier.
Story Protocol: Focusing on intellectual property management on the blockchain, it is more of a decentralized application rather than a physical infrastructure, and may be marginalized in the DePIN category ( Story Protocol ). Story Protocol is trying to tokenize the $70 trillion intellectual property (IP) market. In the AI era, proprietary IP is used to train AI models, leading to copyright infringement claims and large-scale lawsuits, such as the previous litigation dispute between the New York Times and OpenAI. By bringing IP to the chain, Story will enable companies to use their IP for AI model training while allowing any individual to invest, trade, and earn IP royalties. Story has already brought songs from Justin Bieber and BTS to the chain, and launched an IP-centric blockchain and token in February.
4. DePIN track revenue ranking in the past thirty days
The best performing DePIN projects on Solana over the past 30 days
5. Industry event tracking
Roam, an online network service that is essential for global Web3 users, has 2.8 million nodes worldwide, allowing users to achieve seamless cross-border roaming at 30% of the cost of traditional operators. Roam plans to launch a similar incentive mechanism in the second half of 2025, and the spatiotemporal data collected by distributed nodes will become the fuel for training vertical AI models.
Phoenix’s collaboration with TandemAI and Origin Quantum to advance the integration of AI with decentralized physical infrastructure is helping Phoenix take a leading position in the DePIN-AI space.
IoTeX has launched Get Goated Season 2, which involves token rewards and claiming process. The claiming window for $IOTX has closed on March 27, and unclaimed tokens will enter the IoTeX Treasury pool. Sponsors include Geodnet, Uprock, Drop Wireless, and Network 3. The claiming window will start on April 7, and the review period will be from March 28 to March 31, using zkPass verification. This move may enhance community participation and attract more users to participate in the IoTeX ecosystem.
According to the Helium Q4 report released by Messari, Helium network operation data has grown significantly, with operator data offload increasing by 555% month-on-month to 576 TB, mobile hotspots increasing by 14% to 24,800, and daily mobile paid traffic increasing by 99%, showing its disruptive potential in the telecommunications industry. At the same time, Helium unified $HNT as the only token through the HIP 138 proposal, optimized the economic model, and announced a partnership with Telefónica to enter the Mexican market, covering 2 million Movistar users. In addition, Helium was included in the top 20 tokens of interest by Grayscale and included in the COIN 50 index by Coinbase, attracting the attention of institutional investors. In terms of smart city applications, the network has been used for flood monitoring and forest fire warnings in the United States. Helium is expanding through the DePIN (decentralized physical infrastructure network) model to consolidate its leadership in the Web3 telecommunications market.
6. Financing information
Filecoins largest DeFi protocol GLIF released $GLF governance tokens and airdropped 94 million tokens, accounting for 9.4% of the total supply. $GLF will expand to new features such as loyalty rewards in the future. GLIF is expanding to the decentralized physical infrastructure network (DePIN) beyond the Filecoin ecosystem. Currently, GLIF has locked more than $102 million on Filecoin and will support more DePIN networks in the future.
Domin Network, a decentralized business network, announced that it has received strategic investment from Animoca Brands, KuCoin Labs, Web3 Labs.club, IBC Group Official, DWF Ventures, Presto, Outlier Ventures, KnightFury, ThreeDAO, Awakening Ventures and AB DAO. Domin Network is a decentralized business network that uses NFT and DePIN Rollup technology to connect software, hardware and consumer behavior data to the chain. It enables users to earn crypto rewards by sharing their consumption data.
Special statement: All articles of DePINone Labs are for information and knowledge purposes only and do not constitute any investment advice.
This report is produced by DePINone Labs. Please contact us for reprinting.
