You are currently on the new version of our website. Access the old version .
  • This is an early access version, the complete PDF, HTML, and XML versions will be available soon.
  • Article
  • Open Access

21 January 2026

Guarded Swarms: Building Trusted Autonomy Through Digital Intelligence and Physical Safeguards †

,
and
1
Institute for Software Technology, University of the Bundeswehr Munich, 85579 Neubiberg, Germany
2
Department of Computer Science, Sapienza University of Rome, 00161 Rome, Italy
3
Software and Knowledge Engineering Laboratory, University of Molise, 86100 Campobasso, Italy
*
Author to whom correspondence should be addressed.
This article belongs to the Special Issue Intelligent Agents and Their Application

Abstract

Autonomous UAV/UGV swarms increasingly operate in contested environments where purely digital control architectures are vulnerable to cyber compromise, communication denial, and timing faults. This paper presents Guarded Swarms, a hybrid framework that combines digital coordination with hardware-level analog safety enforcement. The architecture builds on Topic-Based Communication Space Petri Nets (TB-CSPN) for structured multi-agent coordination, extending this digital foundation with independent analog guard channels—thrust clamps, attitude limiters, proximity sensors, and emergency stops—that operate in parallel at the actuator interface. Each channel can unilaterally veto unsafe commands within microseconds, independently of software state. The digital–analog interface is formalized via timing contracts that specify sensor-consistency windows and actuation latency bounds. A two-robot case study demonstrates token-based arbitration at the digital level and OR-style inhibition at the analog level. The framework ensures local safety deterministically while maintaining global coordination as a best-effort property. This paper presents an architectural contribution establishing design principles and interface contracts. Empirical validation remains future work.

Article Metrics

Citations

Article Access Statistics

Multiple requests from the same IP address are counted as one view.