TechFlow Logo
Login/ Sign up
ETH Gas
Gwei
Fear
gas
Why Do Robots Need Money? Understanding the Next-Generation Machine Economy Driven by DePIN

Why Do Robots Need Money? Understanding the Next-Generation Machine Economy Driven by DePIN

2026.09.04
Share
The growth of the robotics industry itself does not rely on cryptographic technology; the value of crypto assets lies in bridging the gap in cross-entity collaboration.

Author: Thejaswini M A

Translated by: Chopper, Foresight News

Have you ever imagined that robots need money? This does not refer to AI agents calling API services, but to a physical machine with an actual body. Why would such a device require a cryptocurrency wallet or an independent identity credential?

Many people might find this idea bizarre. However, over a dozen crypto companies are currently rolling out related products, so let's explore the underlying business logic.

Starship Technologies operates approximately 3,000 six-wheeled delivery robots across 8 countries, primarily deployed in major European city centers, with cumulative commercial delivery orders exceeding 10 million.

Serve Robotics, listed on Nasdaq, has expanded its operations to 44 cities including Los Angeles, Chicago, Atlanta, Miami, and Dallas, handling food delivery orders for Uber Eats and DoorDash.

These wheeled box-like devices can carry about 20 kilograms of cargo per trip, traveling several miles per run. If you live in any of the aforementioned cities, you've likely encountered them on the road.

Imagine purchasing a delivery robot to manage a small-scale delivery operation within your community. Running for 12 hours a day, it serves as cheap labor for you, and full ownership remains yours upon purchase.

Referencing Starship's robotic specs, this device can operate for up to 18 hours on a full charge. Eventually, it will need to seek charging stations to replenish power. Starship's solution is to have their robots operate outdoors and return to brand-owned dedicated charging stations. This model works for Starship — the company holds thousands of robots and has the capital to install supporting charging infrastructure across all operational areas.

But if you only own one robot, it must rely on third-party charging stations. Operators could roll out charging facilities just like EV chargers, roadside shops could offer charging services, or other delivery providers could open idle charging ports. With battery capacity comparable to e-bikes, each charge consumes very little electricity. Charging costs range from a few cents to under $1; let's estimate it at 40 cents.

Under the current traditional financial system, how should this fee be settled?

The most straightforward approach is to link a payment card to the robot for tap-to-pay transactions. Bank card payments involve two fees: a percentage of the transaction amount plus a fixed fee of 10–30 cents. Regardless of whether the transaction is 40 cents or $400, the fixed fee remains unchanged because transaction processing costs are constant. For a 40-cent order, the 30-cent fixed fee accounts for a staggering 75%.

As the business scales, the problem intensifies. Fixed fees are charged per transaction. If machines complete $1 billion in low-value service transactions annually in the future, with an average payment of about 32 cents per transaction, the total annual volume would reach 3 billion transactions. At 30 cents per transaction, handling 3 billion trades means fee costs would skyrocket to $900 million against $1 billion in revenue.

Bank wire transfers are even more costly. A cross-border SWIFT transfer incurs fees of $15–50, with each intermediary clearing house charging another $10–30, plus foreign exchange costs. Such payment methods are only suitable for large transfers over $5,000.

The current fiat payment system is designed for a small number of high-value transactions. Machine-initiated autonomous payments are precisely the opposite. If exorbitant fees are the first barrier, the second core contradiction lies in the fact that robots themselves lack independent customer identities.

Even if the fee issue is resolved, charging stations still cannot directly deduct funds from a robot's account. Payment accounts belong to natural persons or enterprises equipped with complete dispute resolution mechanisms. Essentially, you still bear this payment, while the robot merely triggers the swipe action. This is how all global robot payments are currently implemented.

Within the mere two seconds required to authorize an unknown device to draw power, the charging station must verify multiple data points. Is this machine a genuine physical device rather than a script-generated fake request trying to exploit free electricity? Does the device have a good operational track record, or might it abandon the charge midway? If equipment failure causes damage, is there corresponding proof for liability attribution? Granting robots independent identities answers these three verification questions without requiring paper contracts.

Currently, these issues are handled manually. They will invoice you monthly, and if the robot damages anything, they will come after you and sue you. If it were just you and one robot, this system might work. But in a world where two thousand operators and forty thousand machines intersect on the streets, the system requires records so strangers can quickly review them and decide whether to provide service to this machine.

This is the core concept behind the entire solution.

Before diving deeper, we must first clarify a fact: the vast majority of robots will never need this autonomous payment system.

A robot only needs an independent asset account if it simultaneously meets the following four conditions:

  • Devices belong to different owners
  • Business order volumes are small, with random and scattered transaction scenarios
  • No centralized platform handles settlement between both parties
  • Service authorization must be completed on-site within seconds

Amazon has deployed over a million self-developed robots across more than 300 warehouses, all using proprietary scheduling systems, failing the first criterion; platforms like Uber Eats fail the third; and equipment that has signed advance cooperation agreements with merchants fails the second and fourth criteria. This also means this track belongs to a niche market segment, not a universal demand for the entire robotics industry.

Tesla factories currently deploy around 1,000 Optimus humanoid robots, and during earnings calls, officials stated the devices are still in the learning and data collection phase, temporarily not open for commercial service.

Global leading industrial robot manufacturer Fanuc sells hardware products accompanied by the FIELD equipment monitoring platform for fault warnings; this system does not involve fund circulation.

Chinese company Unitree Robotics has lowered the procurement cost of humanoid robots to an affordable range for the masses. The company completed its IPO in Shanghai in August 2026, raising approximately $619 million, with humanoid robot shipments exceeding 5,500 units the previous year. According to the company, buyers use them in diverse real-world scenarios. In May 2026, Unitree launched UniStore, an app store tailored for humanoid robots.

These industry giants are all replicating Apple's business model, keeping scheduling collaboration and payment settlement entirely closed within their proprietary systems, retaining profits on the platform.

Open protocol ecosystems will only survive in the gaps between these major closed systems, such as cross-brand collaboration among robots from different operators and public charging services. The DePIN track market defined by CoinGecko is much smaller compared to the overall robotics industry, and all subsequent analysis must be viewed based on this premise.

First, Know Where You Are

For robots to complete autonomous payments, they must first determine their location. Standard GPS positioning has an error margin of several meters. Starship's CEO has publicly stated that standard GPS precision cannot meet operational needs, requiring navigation accuracy down to the inch level.

Reducing a meter-level error to inches requires correction signals. Ground base stations acquire precise coordinates, calculate positioning deviations, and push calibration data to surrounding devices. This technology is Real-Time Kinematic GNSS (RTK-GPS). The effective coverage radius of correction signals is approximately 30 kilometers, requiring a network to deploy numerous base stations.

GEODNET incentivizes users to mount positioning base stations on rooftops through a token incentive model. Currently deployed globally across 21,000+ devices covering 150+ countries, generating ~$11 million in annual recurring revenue. Multicoin Capital led an $8 million token acquisition deal.

Project revenue stems from providing centimeter-level RTK positioning services to independent devices like delivery robots, drones, and agricultural machinery. At the consumer level, this business actually doesn't heavily rely on crypto tokens; traditional subscription payment models can also be implemented. However, at the infrastructure level, GEODNET has proven that token incentives can rapidly build a global physical network. Relying on traditional enterprises to build it themselves would cost billions of dollars and take decades.

GEODNET Token Supply

Second, Be Able to Interact with Anything

At present, robots produced by different manufacturers run their own independent software systems, making inter-device communication impossible. When enterprises procure hardware, they can only choose a single brand. Hybrid deployment of multi-brand robots requires custom-developed integration programs that currently do not exist.

OpenMind secured $20 million in funding led by Pantera Ventures. Founder Jan Liphardt is a Stanford University professor. The team is developing an open-source general-purpose operating system, OM1, and a scheduling/collaboration layer called FABRIC. Just as Android allows different phones to run the same apps, OpenMind builds a unified software foundation for all robots.

Developers write upper-layer business logic, and the program can simultaneously adapt to Unitree humanoid robots, quadrupeds, and wheeled delivery devices. The ultimate goal is to rely on this universal interaction language to achieve cross-brand device recognition, collaborative operations, and automated payment settlement.

The project builds a machine identity system based on the ERC-7777 standard, defining device behavioral boundaries. If a robot is set as an "auxiliary service" device, the program will automatically reject tasks violating positioning rules. Robots can also cross-verify sensor data with each other to prevent safety incidents caused by abnormal perception in a single device.

OpenMind partnered with Circle to implement zero-gas USDC micro-payments based on the x402 standard, solving the pain point of excessively high fixed fees for small transactions. In an official demo video, the robot successfully made an autonomous electricity payment. The demo ran on a testnet, with no actual on-chain transaction records yet. But this experiment proves that physical devices can independently host wallets, identify purchasable physical resources, and fully automate the entire transaction process. Collaborative interoperability does not require establishing trust.

Third, Possess a Verifiable Identity

Since 2017, IoTeX has focused on blockchain solutions for physical devices, launching two core products to solve identity issues.

Hardware Identity ioID embeds a cryptographic fingerprint into physical devices, allowing them to independently sign and notarize operational actions; Proof of Physical Work W3bstream converts execution actions from the physical world into digitally verifiable credentials on the blockchain.

While IoTeX solved the identity and proof problems, it overlooked credit and financing systems, which is exactly where peaq comes into play.

Fourth, the peaq Protocol

If robots need to make payments, verify qualifications, and retain transaction records for unfamiliar third parties, they require a trust and settlement infrastructure equivalent to human society, functioning similarly to corporate registration systems or the SWIFT cross-border clearing network.

peaq has built an entire supporting system, fundamentally divided into four modules. peaqID serves as a registration credential. Machine NFTs act as ownership records and can be fractionalized using the ERC-3643 standard. ERC-3643 is a token standard that only allows transfers between approved holders. In August 2026, peaq introduced support for P-256 chip signatures, shifting the verification process to hardware security chips.

peaq's machine credit scoring system provides ratings from 0 to 100 based on robot revenue data, activity levels, and contract fulfillment reliability, adopting Moody's-like rating standards ranging from AAA to unrated.

When a robot pays a 40-cent charging fee, the charging station provider can designate the receiving network as Solana or Ethereum. peaq's core value lies in enabling robots to connect to any clearing channel specified by the service provider.

In May 2026, when peaq demonstrated autonomous payment for Serve delivery robots, the funds were ultimately settled on the Solana public chain, not peaq's own blockchain.

Throughout 2026, peaq plays more of an ecosystem integrator role. From January to late August, the project completed 49 development milestones and deployed 20 ecosystem integrations: connecting to GEODNET positioning services, NAVER Maps navigation, and World ID for human-machine identity separation; integrating compute resources from Akash, Acurast, and Arcium; and incorporating Unitree humanoid robots and LG CLOi commercial service robots as hardware terminals.

Corporate registration is backed by national regulatory agencies, courts have verification authority, and banks use SWIFT because the entire industry has standardized transaction formats. For peaq to break free from the constraints of traditional compliance systems, it must proactively engage with Dubai regulators and apply for official licenses. Until compliant credentials are obtained, financial institutions will not recognize this machine credit scoring system as legally binding.

peaq, in partnership with CoinList, launched Initial Machine Offerings (IMOs), allowing users to purchase robot revenue shares. The asset structure was designed by DualMintRWA. A benchmark project is a tokenized vertical farm in Hong Kong, with 80% of its workflow automated. Subsequently, 20 tokenized claw machines were launched. To date, this farm has already distributed approximately $3,600 in returns to token holders.

Even if robots possess positioning systems, operating systems, identity credentials, and credit ratings, the hardware itself still requires capital for procurement. One possibility currently suggested by the market is: the buyer could be an AI agent. Virtuals has already connected approximately 17,000 on-chain AI agents to Solana's BitRobot network. AI agents pay to hire physical robots to complete offline tasks, with funds held in smart contracts and automatically settled upon task confirmation.

Software agents paying robots versus robots transferring funds to each other represent two different models, with the former offering stronger commercial sustainability. AI agents possess digital capital, business objectives, and computing power, but cannot intervene in the physical world; physical robots have mobile hardware capabilities but lack native funding and business needs. When robots pay each other, the devices often belong to the same enterprise, making internal bookkeeping far simpler and more efficient than actual on-chain transfers.

Looking at the broader market, within the overall market cap of coins in CoinGecko's robotics track, GEODNET stands at approximately $100 million, while peaq is around $55 million.

It is clear that this track remains a niche within a niche field.

Industry statistics show that, according to the IFR's "2025 Global Robot Report", 542,000 new industrial robots were installed worldwide in 2024. The operational stock totals approximately 4.66 million units, with over 2 million deployed within China.

In an external research report published by JPMorgan in July 2026, it was projected: global robot market sales reached approximately $100 billion in 2025; under a baseline scenario, annual market sales will hit $2.5 trillion by 2035; a pessimistic forecast stands at $500 billion, while an optimistic outlook reaches $8 trillion. The same report calculates that the humanoid robot market will grow from $2 billion in 2025 to $300 billion in 2035 under the baseline scenario.

This is the target market the crypto industry aims to penetrate and build payment and identity foundations for. The crypto industry has never lacked grand visions, and this time is no exception.

The concept of autonomous machine interaction is not new; similar ideas were proposed early on in the IoT track. As far back as 2015, IBM and Samsung demonstrated a washing machine capable of autonomously purchasing laundry detergent via Ethereum, named the ADEPT project. Months later, IBM invested $3 billion into its IoT business. However, the IoT sector eventually pivoted toward data monitoring. Billions of devices merely uploaded operational data to vendor backends. Although the entire technology stack was fully deployed and devices obtained vendor server communication certificates, automatic transactions between machines never achieved market scale, and the traditional payment system remained untransformed.

Robots differ fundamentally from ordinary IoT devices. Climate control equipment costs only $200 to procure and operates in a fixed manner. In contrast, robots entail high acquisition costs and possess the ability to generate revenue. Once devices can generate income, credit and insurance become absolute necessities, requiring third parties to verify their contract fulfillment and payment capabilities. For years, IOTA has aimed to build a public chain for machine economies, but its business direction subsequently shifted; its implementation cases focus on government notarization scenarios such as Kenyan customs documents, UK port trade certificates, and Argentine organ donation registries. Government agencies procure blockchain identity tools for real data recording, but the M2M automated payment business has consistently failed to breach the market.

The growth of the robotics industry itself does not rely on cryptography. The value of crypto assets lies in filling the gap in cross-entity collaboration: granting devices publicly verifiable identities, self-custodied encrypted wallets, and low-cost micro-payment channels.

Let us envision an open robotics labor market where anyone can rent a robot they do not own to complete tasks. This market requires comprehensive supporting infrastructure: verifiable device identities, low-cost micro-payment solutions, and performance credit archives to avoid renting out devices plagued by battery or malfunction issues; it should also provide rental value-added services such as high-precision positioning, cloud computing power, manual remote control, and idle charging stations, which DePIN is perfectly suited to handle.

Of course, if Tesla, Amazon, and leading domestic hardware manufacturers continue to keep the entire business closed within their proprietary ecosystems, an open market may never truly materialize. Once the open market lands, new business questions will immediately follow: who will provide financing services for robots capable of generating on-chain revenue? Who will underwrite the equipment? How will robot clusters be used as collateral? Who will build the matching platform to receive warehouse-published orders and accept bidding quotes from robots? After the identity system is established, asset securitization becomes the next phase's proposition. Tokenization only truly releases its value when assets can circulate freely.

DePIN is merely a lightweight underlying channel serving non-closed-loop operational niche scenarios within the robotics industry. Industry giants have no obligation to drive traffic/business to this ecosystem.