The "Green Blockchain" project aims to develop a high-impact demonstrator showcasing the benefits of blockchains for climate preservation.
This blockchain is 'green' in two ways:
The project leverages existing technology to be extended/adapted for climate data use. It includes an introduction to blockchains and details on the available technology (Section 2), the use case and project goals (Section 3).
Currently referred to as distributed ledgers, blockchains often face reputation issues, particularly regarding environmental impact. Originally developed for cryptocurrencies, they were associated with dubious markets and illegal activities. Bitcoin's energy consumption further tainted this image, with a single transaction equating to a four-person household's weekly energy use. The saying goes, "Burning the planet to launder money" isn't appealing. However, blockchains have positive aspects too. For instance, consider online payments: normally, a bank validates them. But imagine being in Congo, wanting to buy a house without a notary for validation. A digital wallet on your phone, containing money and a way to validate your purchase and change of ownership without a notary or official land registry, would be quite convenient.
Blockchain indeed networks participants who may not necessarily trust each other, enabling them to share a trustable ledger* – coherent, immutable, and verifiable by all without a central control authority. Figure 1 illustrates a generic process of validating and constructing a blockchain-type ledger.
Figure 1 - A user provides data using his blockchain client (1), the blockchain client encapsulate data in a signed transaction and submits to the blockchain network (2), a validator group in the blockchain verifies transactions in a consensual way to add them to the register (3).
Thanks to its properties of "disintermediated" validation and programmability through smart contracts, blockchain can be used for a variety of beneficial applications: transparency and traceability of the food we eat, traceability of green energy, self-consumption, and neighborly energy markets, as well as climate monitoring – as evidenced by many projects and startups today¹.
*A transaction ledger, data, codes, etc.
Committed to more frugal digital technology, the CEA has helped make blockchains greener. But what does that mean? It means that the CEA has replaced the energy-consuming element in Bitcoin, the proof of work, with an algorithm based on validator committees. To understand this better, one needs to delve into the technical details of how the ledger is constructed and validated by the network.
Whenever an actor wants to alter the shared ledger, such as adding new data, it's necessary to verify the data's validity to adhere to the ledger's rules. For instance, in land registries, a rule could be that each property has only one owner. An actor proposes a registry update to the network, which then validates or rejects the change. Ensuring every network node works on the same ledger version is a challenge in distributed networks, where nodes communicate over unreliable channels. Scientists refer to this challenge as the Consensus problem in distributed networks. Over 30 years ago, MIT researchers established that solving the Consensus problem is impossible if nodes can fail and communication delays are unbounded. To overcome this, Consensus algorithms assume a known number of nodes and a maximum number of fail-prone nodes, assuming temporary communication breakdowns and known recovery times. This problem is exacerbated in blockchain networks where the number of nodes is unknown. Bitcoin addressed this with the proof of work mechanism, where nodes solve a cryptographic puzzle, akin to finding an unknown key to open a door. The key, in more technical terms, is an alphanumeric sequence, and opening a door involves using a hash function that takes in data (d) and the key (k), outputting another alphanumeric sequence (o). If o starts with a certain number of zeros, then k is the correct key; otherwise, the operation is repeated. Finding the key depends on the computational power of the nodes. The faster they perform the hashing operation with a new key, the higher their chances of finding the correct key compared to slower nodes. To add new data to the blockchain, the first node to solve the proof of work writes the ledger, updating its copy and sending it to others*. The validator is rewarded for this costly and time-consuming task. This reward motivates validators to follow the ledger's rules; otherwise, their copy is rejected and they receive no reward. The CEA has proposed alternative consensus algorithms to proof of work, based on new ways of selecting validator nodes. These nodes are chosen by a committee based on known and verifiable criteria. The elected committee reviews the registry update proposal and makes a decision through voting rounds. The committee's decision, to be valid, must be signed by a majority quorum and then disseminated to other network members. In these algorithms, blockchain network consumption depends only on the number of validator nodes, unlike Bitcoin, where it's linked to the currency's valuation. Recent studies² show that these algorithms can significantly reduce energy consumption compared to Bitcoin, even lower than the centralized Visa system. Notably, the CEA designed the consensus protocol for the Tezos³ blockchain.
Figure 2 - shows the energy consumption per transaction in blockchains (Polkadot, Cardano, Tezos, Algorand, Hedera) that use consensus mechanisms without proof of work, compared to Bitcoin and Visa. The energy consumed is inversely proportional to the throughput.
*The security of this mechanism relies on the unpredictability and the impossibility of pre-solving the cryptographic puzzle. Therefore, it's impossible to predict which node will be next to write in the ledger, preventing malicious actors from targeting it.
Since 2017, CEA has invested in blockchain technology and helped develop applications that have a positive impact on society. Connecting Food is arguably the most emblematic example: CEA co-invented an online audit application for food supply chains with Connecting Food, ensuring that products comply with brand criteria and labels (non-GMO, organic, origin in France, etc.). Connecting Food, now in the scale-up phase and present in the European and American markets, received the European EIT Digital Award for the best Deep Tech technology in 2021⁴.
Other applications have been co-developed by CEA in the energy and manufacturing sector in collaboration with startups and major companies (such as EDF R&D, Veolia, Dassault Aviation, Engie, and Bureau Veritas).
In the field of applications, CEA focuses on the following technical challenges:
Trustworthy smart contracts: CEA assists its clients in implementing their applications through dedicated and verified smart contracts. Smart contracts are computer codes that run on the blockchain network. For this reason, the codes are immutable and execution is integrated. Smart contracts allow specific validation rules for the application to be encoded in the blockchain (e.g., product validation rules, maximum allowed quantity rules for an asset, etc.), as well as managing the lifecycle of digital assets (creation, ownership transfer, escrow, destruction). CEA has developed design and validation tools to ensure that smart contracts are well-coded and do not contain bugs or vulnerabilities.
Securing oracles: From an architectural perspective, the blockchain can be considered as a component within a larger software system. In cases where the blockchain is used as a distributed database for purposes beyond digital assets, applications built on a blockchain may need to interact with external systems. Thus, the validation of transactions on a blockchain can depend on the states of external systems. For example, a payment may be conditional on the physical properties of a product/service or the completion of a real-world activity such as shipping a product. Generally, this data is provided or certified by third parties. Presenting the state of external systems within the blockchain execution environment is done through a third-party service called an oracle. An oracle queries and verifies external data sources, then relays this data into the blockchain in the form of a transaction (Figure 3).
Figure 3 - Blockchain architecture with oracles.
Securing the link between the external source, the oracle, and the blockchain is paramount. There are two types of oracles: centralized and decentralized. A centralized oracle introduces a trusted third party into the system, which can become a single point of failure. Alternatively, a decentralized oracle is designed based on multiple independent services and data sources, overcoming the single point of failure. A decentralized oracle can use multisignature or voting to achieve consensus on the validity of injected data. Similarly, at the hardware level, redundant IoT devices could be deployed to deal with malfunctioning IoT devices that produce erroneous/malicious data. Subsequently, a smart contract can verify the consistency of data collected by different IoT devices. CEA has developed hardware and software components to secure the connection between a client device (IoT, embedded device, Cyber Physical System) and the smart contracts deployed on a blockchain.
When IoT devices interact in a decentralized or distributed topology, each becomes the trust anchor for the actions it performs. This requires anchoring security in the farthest, smallest, and constrained objects to ensure end-to-end security by design between the data-collecting object and the smart contract. A smart contract library has been developed for this purpose, enabling connections with (or between) remote objects that incorporate hardware security components. Technological development in this field involves innovation and R&D since blockchain technologies were not originally designed to operate in a low-power consumption context.
As a distributed system, every blockchain is subject to the theoretical constraints of distributed systems: partial observability, difficulty in ordering all events over time, impossibility of consensus... However, the non-energy-consuming alternative algorithms developed by CEA enable consensus to be achieved under the assumptions presented in section 2. They are a key element of trust in the blockchain. If an actor exhibits behavior that deviates from what is expected by the consensus algorithm (due to a failure or malicious intent), it is crucial to identify and provide evidence of the behavioral deviation, if possible, before the validation of the next block. Techniques such as formal specification of distributed systems and multi-trace compliance are currently being explored for this purpose. The principle involves recording the messages sent and received by each actor and evaluating, through symbolic execution, whether there is an execution path compatible with the message history within the consensus algorithm. These very general techniques can also be used to enhance trust in other elements: at the level of smart contracts, in the code of oracles, and at the application layer built on the blockchain.
CEA has developed consensus algorithms for green blockchains, formal methods, and validation tools to ensure the correctness of consensus algorithms and smart contracts. They have also created software and hardware technological components to certify oracles and secure connections with IoT. CEA also has experience in implementing blockchains for industrial applications.
This project aims to bring together the technical components and expertise that CEA has developed in recent years in the areas of non-energy-consuming consensus algorithms, formal validation methods for distributed systems, and secure components interacting with IoT. Indeed, these competencies have been developed by separate teams (LIST and LETI), and we believe that the time has come to consolidate the work of these teams around a single high-value, disruptive technological objective, namely: A low-consumption, trustworthy, and securely interacting blockchain with sustainability in the physical world.
To effectively challenge our blockchain "greening" project, we have selected a "flagship" use case with a strong societal impact: using blockchains as a tool for monitoring greenhouse gases to safeguard the climate.
Use case: Low-cost and trustworthy processing and sharing of climate data from heterogeneous data sources.
The use case provided by the Laboratory of Climate and Environmental Sciences is detailed in the following. Our project will enable the development of a digital, green, and open-source infrastructure aimed at sharing a trustworthy climate data registry along with data processing algorithms in the form of smart contracts.
The case study is linked to ICOS, the European distributed research infrastructure measuring greenhouse gases*. ICOS's mission is to quantify atmospheric concentrations of greenhouse gases (GHGs) and land and ocean fluxes in Europe by producing standardized observations to better understand the carbon cycle in the context of climate change. ICOS data also support policy decisions to combat climate change and its impacts. In ICOS, the Atmospheric Thematic Centre (ATC) is operated by LSCE/DRF/CEA. The ATC has two main missions: one involves instrument metrology, and the other concerns the processing of data from the European network.
ICOS network
The ICOS network was initially deployed as a background network far from cities with a focus on natural greenhouse gas (GHG) flows (biospheric and oceanic). However, in the context of validating the implementation of the Paris Agreement and the growing importance of cities in terms of anthropogenic emissions, urban GHG networks have recently developed with reduced constraints in terms of measurement accuracy due to their proximity to GHG emission sources.
In this new context, ICOS is leading the European H2020 project PAUL (2021-2025), which aims to enhance our understanding of specific needs related to GHG quantification in urban environments. PAUL will develop three pilot GHG measurement observatories in the cities of Paris, Munich, and Zurich, and provide services to support cities in implementing their climate action plans.
As part of the moonshot project, the proposed case study involves collecting, processing, and visualizing time series data of CO2 and associated parameters from 10 low-cost sensors deployed in the Paris region for a two-year period by LSCE. It includes the collection of raw time series data of CO2 and associated parameters from the sensors, using the latest IoT standards as much as possible. Within the blockchain, CO2 calibration will be performed using the associated parameters (pressure, temperature, humidity, etc.) also measured and calibration gas bottles, regularly measured by the network. The blockchain will also perform data averaging for the CO2 time series and automatic quality control.
The case study also includes the development of a dashboard representing data from various instruments to complement the quality control process through automatic outlier detection. This development within the green blockchain is in excellent alignment with the FAIR approach undertaken by ICOS for several years now (H2020 project ENVRI-FAIR). The blockchain developed in this way will serve as an immutable registry of GHG measurements and will enable the tracing of different data flows throughout their lifecycle, from raw measurement to the final product, while allowing free access to the data through user-friendly and informative tools.
*ICOS ERIC, the legal entity, was created in 2015. ICOS now has the ESFRI landmark label, and the French component of ICOS is a TGIR/IR.
The green blockchain provides a reliable and interoperable digital tool that aims to offer the following properties compared to current infrastructures:
Immutability and Verifiability: All collected data and associated processing, when recorded in a blockchain, are immutable and verifiable by everyone. This means that it is possible to create immutable and indisputable shared records. These properties of secure and transparent archiving will prevent any accidental data loss while facilitating their availability/publication. For example, this can be useful when publishing a study/report for the scientific community, the political world, or the public.
Timestamping and Traceability: Climate data, such as greenhouse gas-related data, is collected by sensors. For each sensor, it is natural to record raw measurements as time series data. Processing algorithms filter/cleanse these measurements, generating qualified data. By observing the raw measurements generated by a sensor, it is also possible to detect any malfunction, thus anticipating a calibration operation.
Blockchains inherently timestamp the data and the computational steps performed on the data. Calibrations (per sensor) are also timestamped operations recorded in the blockchain. This allows for tracking and tracing the entire history of a time series and revisiting relevant measurements/data qualifications/calibrations.
Scalability and Flexibility: Based on a technological framework built on blockchain principles, it is extremely easy to add actors to the system who could bring new sensors, data, algorithms, and calibration techniques. To add a new sensor, you would simply register it on the platform (e.g., via a mobile application) and connect it as an oracle to an existing smart contract for data registration/qualification. A monitoring smart contract could also be associated with the registered oracle. Such a smart contract could trigger an alert when measurements degrade, indicating that a calibration operation is needed. Other actors could also add new data processing algorithms or monitoring through new smart contracts, which would be immediately available within the community. This flexibility (connecting a new component to existing ones) and scalability (adding more advanced components) would facilitate the creation of a collaborative knowledge-sharing ecosystem.
Note: The link with AI: It is important to emphasize that in the long term, these distributed and reliable registries could form a solid foundation for AI algorithms. AI algorithms could enhance/optimize data qualification smart contracts, for example, but also predict and optimize sensor maintenance.