|
Blockchain and IoT
The multiplication of IoT devices means that the data of billions of devices needs to be managed. IoT devices also need to communicate with each other in a secure way.
The report points out that the client-server-based approach, where one centralised system handles the data and communication, is reaching its limits. At least two major problems were identified: there is a single point of failure and a risk of vendor lock-in (lack of interoperability).
The report explains that the centralised solution offered by a single vendor could be replaced by a decentralised IoT platform developed and run by a consortium of interested stakeholders. Thus, an IoT device would package its data together with metadata like the device ID and a timestamp, hash the data and electronically sign it with its private key, and then send it to the blockchain. This way the data would be sealed (by the hash) and uniquely identifiable and findable (through the public key of the data record). This approach would avoid the single point of failure. It would also facilitate machine-to-machine communications as blockchain would provide a common trust communications layer between devices of different types and manufacture, hence supporting interoperability.
Other advantages of this new approach:
-It would make it easier for IoT devices to verify the authenticity of instructions being sent to them.
-Blockchain and IoT work well together to provide verifiable identities to natural persons and objects.
-Smart transactions between machines would be facilitated.
Blockchain and AI
The report notes that data is concentrated in a few hands. There is also a lack of data-sharing standards, so data often remains in silos. Moreover, it results difficult for companies and individuals to control the use of their data once they have sold it. On the other side, AI developers might struggle with data quality.
Blockchain could support the development of decentralised, open markets for AI training data. It could be used to identify and record individual data points and small data sets, making it possible for information owners to package the data they generate. For larger data sets, blockchain provides a record of their provenance and secure them against tampering. The report mentions the Ocean Protocol, which supports this approach.
In addition, blockchain could be used to facilitate collaboration among AI developers. For instance, it could support decentralised platforms for the development and dissemination of AI models, tools and services. The report mentions here the project SingularityNET.
Blockchain could also be deployed to provide access to trained AI models for individuals and companies. This could be extremely useful for a company having data but no AI expertise.
In all these cases, blockchain provides trust and security in the data.
On the other hand, AI can support blockchain, for instance to increase security by running plausibility tests and detecting anomalies before data enters the chain. AI could also detect attacks on a blockchain.
Convergence of blockchain, AI and IoT: the example of smart cities
The report mentions the example of smart cities as an example of how blockchain, IoT and AI can converge. Blockchain could serve as the intermediary identity and authorisation layer between IoT devices (such as sensors) and AI-assisted infrastructure administration systems. It could also serve as a secure data storage or data coordination platform for heterogeneous data sourcing.
Data registered in the blockchain is usually very well structured, which would facilitate its use by AI in real time to support decisions about city activities. For instance, sensors, AI and blockchain used together could improve public transport.
Challenges and risks
-Technological: Blockchain faces scaling and interoperability challenges. The large-scale convergent platforms for blockchain, IoT and AI will need to be interoperable. Furthermore, these platforms will deal with serious security challenges.
-Legal and regulatory: compliance with GDPR must be ensured and liability in blockchain-based platforms is a big challenge.
-Governance and ethics: privacy issues. For instance if the movement of every citizen is recorded to improve transportation and traffic.
Recommendations
1) Ensure adequate and targeted funding: Europe should fund research on existing convergence use cases.
2) Promote best practice and responsible ownership of these technologies.
3) Consider any adaptation of regulatory processes to meet the demands of these new types of platforms.
4) Promote private sector engagement and public/private partnerships.
5) Provide regulatory clarity.
6) Do not regulate too early.
7) Consider ethical issues.
For any question on this issue, do not hesitate to contact Camille Dornier: camille.dornier@eurosmart.com
|
