Software Specification and Automatic Code Generation to Realize Homeostatic Adaptation in Unmanned Spacecraft

Abstract

Long-term space missions are requiring more and more autonomy from Unmanned Space Vehicles (USVs). USV control architectures are required to support effective adaptation to unpredictable internal and external changes. Thus, continuous assurance of efficient management of resources is essential to realize such requirements. Inspiration on the self-regulation provided by the physiology of living systems is an attractive solution to achieve continuous adaptation to the environment by changing internal conditions. Physiological functions provided by nervous system reflexes at a very low control level are the foundations for self-regulatory mechanisms such as homeostasis. USVs can get highly-viable and ultra-stable by building artificial self-regulation in order to support very long missions. This paper presents aspects on how to endow USVs with Artificial Nervous Reflexes (ANRs) to the USV control architecture by means of applying physiological principles of self-regulation for operational resilience and persistence. The above homeostatic concept is applied to a composite orbiter (i.e. a USV) for the BepiColombo mission to Mercury. The ANRs studied are required to guarantee self-regulation of response time (latency), operation temperature (thermoregulation), and power consumption (energy balance).Results from experiments carried out based on the realization (software generation) of the above physiological paradigm from software specification are also presented.