By Andrew J Lewis. Contact: ceo@globalymbe.space
Meet DETENTE, an independent vehicle mission management system able to support, and in some cases sustain, spacecraft decision-making when delay, distance and uncertainty make constant human supervision impractical. Rather than acting as a narrow autopilot, DETENTE is being shaped as a modular mission mind: a software architecture that can weigh conditions, update priorities, learn from outcomes, and recommend or execute bounded decisions in pursuit of mission success.
What makes this especially exciting is that DETENTE is not conceived as a monolithic block of code. Its design is modular by intention. Separate functional layers can represent internal state, environmental perception, motivational pressure, action selection, reinforcement, and policy persistence. That matters in space operations, because real missions do not fail or succeed on one variable alone. They succeed because power, thermal state, timing, risk, payload needs, communications windows, vehicle health and mission objectives are continuously balanced. A modular architecture gives us a way to model that balancing act transparently, test it incrementally, and refine it without rebuilding the whole system each time.
Scientifically, DETENTE draws on a mixture of behavioural reinforcement logic, affective state modelling, and structured policy adaptation. In practical terms, the system is being tested not only for whether it “does something”, but whether it changes in a disciplined way after experience. Current test outputs focus on learning events, valence shifts, and policy evolution across persistence chains. In particular, the project incudes concepts modelled from Aristotle, Jeremy Bentham, Sigmund Freud, Plutchik et al and BF Skinner.
That is why the recent test runs have been so encouraging. They suggest DETENTE is already becoming more than a scripted rules engine. The architecture is showing the kind of structured internal change one would hope to see in a future mission-management layer: not consciousness, not mystique, but an orderly capacity to register conditions, select among pressures, and alter later behaviour in light of earlier results. For a deep-space vehicle, that opens an important prospect. A system of this kind could one day help arbitrate between competing demands such as conserving energy, preserving vehicle viability, maintaining timetable discipline, or delaying a manoeuvre when system stress is rising.
The space application is therefore not speculative ornament. It is a natural destination for the design philosophy. A vehicle in transit to the outer Solar System, or operating at great light-time distance, needs more than sensors and actuators. It needs an internal executive layer able to interpret mission context and remain robust when conditions are imperfect. DETENTE is being developed with precisely that sort of bounded autonomy in mind.
Just as important, DETENTE is not the end of the story. It is also feeding into COLLEGIUM, the broader multi-agent research environment now being prepared around heterogeneous agents with differing traits, action biases and social roles. The project aim there is to move beyond isolated behaviour into interaction, reciprocity, cooperation, conflict and emergent order. In that sense, DETENTE is both an operational mission concept in its own right and a foundational module inside a larger experimental framework.
That combination is what gives the programme its momentum. We are not simply building a control routine. We are exploring whether a spacecraft can be given a disciplined internal decision architecture — one that is modular, testable, extensible and ultimately capable of supporting the next generation of autonomous mission operations.
Andrew J Lewis Msc FLS MIET
Chief Future Officer. Globalymbe Limited Contact: ceo@globalymbe.space

About Globalymbe Limited:
Globalymbe Limited is a UK micro-SME developing a distinctive portfolio of applied intelligence systems across strategic risk, environmental insight, behavioural simulation, and advanced decision support. Its domains span geopolitics and peace-building, climate and infrastructure analytics, energy and commercial assurance, and experimental multi-agent software for mission management and organisational reasoning.
The company’s ambition is to turn rigorous interdisciplinary thinking into practical, market-ready tools that help institutions and enterprises anticipate change, test scenarios, and make better long-range decisions under uncertainty.