Answers are grounded in this paper and cite their sources.
Adaptation of the Hyperledger blockchain model of electronic voting to emergent adaptive neural networks and risk management in cyber-physical systems. The article presents the concept of combining a decentralized blockchain voting model (on the example of Hyperledger) with emergent adaptive neural networks for dynamic risk management in cyber-physical systems (CPS). The features of the functioning of distributed ledger platforms (permissioned blockchains), their consensus mechanism and the potential of using smart contracts to automate collective decision-making processes are considered. It is shown how the e-voting model can provide transparent and reliable coordination of updates or actions in multi-agent neural networks, as well as increase the security and accuracy of risk management in the CFC. The article presents mathematical approaches to formalizing the integration of neural networks with the blockchain level, describes consensus algorithms based on Proof-of-Learning and voting. The mechanisms of decentralized storage and validation of model weight updates, inclusion of risk assessment logic in smart contracts, and the possibility of using tokens to stimulate correct updates and prevent data poisoning are considered in detail. The methods that allow combining the emergent capabilities of collective self-learning of neural networks and the advantages of blockchain technologies (data immutability, automated security policies, audit) are analyzed. Using the example of managing autonomous robotic platforms in production systems, the paper demonstrates how such a synergistic system can increase resilience to internal and external threats and reduce emergency response times. It is proved that the distributed registry acts as an “immune system” during the evolution of neural networks and prevents the legitimization of dangerous changes in model parameters. Promising directions for future research are outlined, including the creation of prototypes of high-loaded CFCs with emergent II modules and the formation of formal reliability criteria for such hybrid solutions. In general, the presented adaptation of the blockchain voting model for the tasks of dynamic risk management and distributed learning is promising for building more flexible, transparent, and secure cyber-physical systems.
No takes yet. Share an insight, caveat, or question.
Prokopovych-Tkachenko et al. (2025) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: