Component-age entropy as a controllable driver of collective order

Full abstract

Read the full abstract

Multicomponent systems, from biological tissues to engineered swarms, are assembled from parts of different ages; whether this heterogeneity degrades collective order or merely marks it is an open causal question spanning aging biology and engineering reliability. We address it with a controlled physical model: a bristlebot (vibrating-robot) swarm that tests whether the entropy of the component-age distribution drives the loss of macroscopic order. Each robot carries seven critical components with independent manufacturing dates; age mismatch is quantified as τ_sick, the normalised Shannon entropy H(τ) = −Σ pᵢ log₂ pᵢ of the component-age distribution. Collective order is measured by a Hierarchy Index (HI). An open simulation (GPLv3) reproduces the core prediction: at τ_sick = 1.00, HI falls below baseline and the swarm becomes chaotic. A falsifiable nine-test matrix (Cohen’s d = 0.3, α = 0.05, β = 0.80, n = 60/group) separates calendar age, component micro-age and age entropy. The low-cost (~€1,000), fully open platform bridges aging biology, engineering reliability and stochastic thermodynamics. Demonstrating that age heterogeneity as a parameter suffices to disrupt collective order would provide a quantitative bridge from an information-theoretic driver to a system-level thermodynamic outcome, with direct relevance to mosaic aging and engineered distributed systems.