Deep space exploration with surface based robotic elements presents a number of significant challenges, not least of which is the lack of real-time, high bandwidth communication between a surface element and mission control. The lack of timely data combined with the unstructured and potentially hazardous nature of the terrain creates a difficult mission planning environment. Given the paucity of up-to-date data it is difficult to construct robust timelines at the task level. As a consequence such missions can be characterised by recurrent safe-modes, significant downtime and reduced science return. Lag in the communications cycle exacerbates this problem by increasing operational workload and delaying return to active science operations. It is worth noting that even in mid 2005 with a reduced and mature approach for the NASA MER rovers, operations costs are in the region of $2.5M/month. A conservative planning approach is often adopted to improve timeline robustness but inevitably science return is lower as a result. The use of greater on-board autonomy is therefore a critical consideration for such a mission. In particular it is essential to determine just what level of autonomy is appropriate, whether a suitably mature technology implementation exists and to validate these conclusions in a representative way. The current suite of Packet Utilisation Services (PUS) as defined in ECCS-E-70-41 [1] provides an appropriate starting point for the development of an operational autonomy concept for the forthcoming ExoMars mission. Although PUS services such as event monitoring, event action coupling and OBCP’s provide the backbone implementation for FDIR and operational autonomy concept, additional functionality is required to meet the requirements of deep space operations. In particular there is a need to accommodate timeline management from a global perspective in response to variability (during execution) and uncertainty (at mission planning time) with regard to on-board resources such as memory and power. For example the actual power or memory used during payload activity will often be different from that predicted by mission planners. In the pessimistic case this will halt the timeline as there will be insufficient resources to carry out the planned tasks which remain. Alternatively execution may result in a relative abundance of resources if estimates were overly conservative. These could of course be used to carry out additional science. Greater on-board autonomy, which extends the level currently supported by existing PUS services, has been proposed as a means of alleviating this inefficiency. Autonomy in this sense is defined as having software element which can modify or create a rover timeline. The motivation being that an on-board element, with access to actual on-board state is