ADOS (AirTrace Decentralized Oracle System): an advanced AI-based oracle system for securing off-chain IoT data integrity when injecting in the blockchain
According to several studies, the adoption of Blockchain (BC) technologies in the realm of IoT (Internet of Things) is going to grow exponentially during the next years, i.e. the global blockchain Internet of Things (IoT) market will become a three billion dollar industry by 2025. Worth $32 million (USD) in 2018, several reports put the industry at a 91% growth rate starting 2021. As an increasing number of businesses invest in IoT to connect and share data, company leaders are recognizing the potential of blockchain technology to ensure the transfer of information is secure.
However, applying blockchain to IoT is a still a big challenge today, as it entails properly facing several hurdles like oracle trustworthiness, scalability, interoperability, resource consumption, privacy, and processing delays, among others. In this project our focus is to develop new approaches for distributed oracles to be tested with AirChain, a blockchain-agnostic platform that allows any blockchain practitioner (ranging from hobbyists to large corporations) to easily integrate blockchain capabilities in their IoT networks.
Our efforts for ADOS aim at developing an advanced solution for a distributed system of oracles in IoT scenarios. IoT sensors present some promising options to improve security, reliability, and authenticity of the information provided by oracles, that we want to explore via cutting-edge technologies like Artificial Intelligence thanks to anomaly detection, for instance, and the application of advanced identity verification procedures for each particular IoT device. Thanks to our developments, the ONTOCHAIN community will have the chance to resort to our solution in terms of oracle handling and thus increase the reliability, security and authenticity of any information injected to the ledger.
Most advanced oracle protocols do not take advance of the intertwining ontological relations of IoT devices in particular scenarios (agriculture, industry, etc.). Resorting to geospatial (or even more) correlations could lead to cutting-edge anomaly detection algorithms to maximize data integrity before injecting into the blockchain.
ADOS will be especially used in scenarios where IoT data integrity must be monitored and quantified before injecting into the blockchain to continuously track it and maximize its reliability for effective auditing.
ADOS will be based on an anomaly detection scheme whereby the algorithm benefits from the ontological relations of IoT devices measuring correlated magnitudes, so that it is possible to not only blockchainize the specific data, but also its integrity factor.
These functionalities will offer a specific integration scheme that will take into consideration specific metadata from IoT devices (space location, time, measuring magnitude, etc.) in order to track the integrity factor of data produced by the IoT device.
There is little to no research on distributed Oracle systems for IoT. If we even go further the line, we do not find, as far as we know, any work using AI in this kind of scenarios. That is why we want to cover this particular gap with an effective solution.
ADOS will offer bidirectional value: B-IoT integrators will find it useful to track and quantify the integrity of data they inject into the blockchain, thus paying extra money that will attract oracle creators that will generate information useful for the anomaly detection scheme.
We think of a real-life or a simulated scenario where we showcase an IoT sensor network whose data is injected into the Blockchain. Thanks to ADOS we will see how data integrity can be measured before it is blockchainized, so that audits track not only data, but its integrity factor as well.