Technology

Fetch.ai, a Cambridge-based artificial intelligence lab building an open-access decentralized machine learning network for smart infrastructure, announced today that the team of the “Economy of Things” (EoT) strategic advance engineering project at Bosch Research will be initiating machine learning trials on the Fetch.ai Collective Learning network. The system will attempt to predict potential failures in Bosch’s machinery while maintaining data privacy. “Secure and trustworthy computation across several participants, while keeping the raw data and possibly even the learned model private is key to unlock the true value of distributed data,” said Dr. Alexander Poddey, leading researcher for digital socio-economy, cryptology, and artificial intelligence in the EoT project. “In our view, collective learning is a key enabler to leading digital socio-economy to efficiency.”

Bosch will now also systematically transfer the self-sovereign identity (SSI) principle to the B2B sector – in other words to business relationships between companies. A further step will involve transferring it to secure and trustworthy interactions between machines. Currently, the implementation framework is being provided by the European data infrastructure project Gaia-X, for example. Before a specific Gaia-X operation can be launched, the central Gaia-X Federation Services (GXFS) will be needed as a basis for establishing ecosystems. For example, the team of the “Economy of Things” strategic advanced engineering project at Bosch Research worked together with other partners as part of the GXFS to draw up the technical specifications for the aspects “Identity and Trust” using SSI. The reference implementation they created is now being used in Catena-X, the implementation initiative for the automotive industry. The basis for this is also being laid in IDunion, where Bosch is part of the team developing a fundamental identity data infrastructure.

Distributed Ledger Technologies (DLT) such as blockchain are very memory and energy intensive when transactions need to be processed on a decentralized basis in real time. For a number of years, Bosch has been working with partners on a second-layer protocol known as “Perun”– a solution that facilitates DLT scalability in line with economic and ecological considerations. Technical university TU Darmstadt and Bosch Research have been collaborating on an initial open-source implementation as part of the “Direct State Transfer” project on GitHub. As the project is increasingly reaching technical maturity, the time has come for the next organizational step. The framework is being developed further as a Hyperledger Labs project under the name “Perun”. Hyperledger is an open-source community that focuses on developing a series of stable frameworks, tools, and libraries for enterprise-grade blockchain deployments.

The team behind the “Economy of Things” strategic advance engineering project at Bosch Research is developing innovative concepts and software solutions for digital master data and certificate management that boost data quality and sovereignty. This includes using the principle of self-sovereign identity (SSI), which enables digital identities to be managed independently without relying on a central identity provider.

Prototypical DLT applications already exist, but the technical requirements for sustainable business models are still missing, for example when it comes to scalability: Scalability is one of the great challenges with DLT, because ideally tens of thousands of transactions per second have to be processed in real time. This is very memory and energy intensive. The team of the strategic advance engineering project “Economy of Things” has set up the open source project “Direct State Transfer” (DST) as a solution to implement the so-called second-layer protocol “Perun”.