AUTONOMOUS DECISION MAKING IN VERY LONG TRAVERSES (ADE)

You are here: Home » Background » Proposed concept

Proposed concept

ADE Demonstration scenarios

In order to better understand the technical choices done in the frame of the ADE proposal preparation, it is very much important to define from the very beginning the demonstration scenario in which the overall system will be nominally operating and validated.

The ADE demonstration scenario is defined with the goals and ambition of:

  1. Fulfilling the objectives of the OG10 call
  2. In particular, since the focus of the ADE project is to prove a system with autonomous deliberative capabilities, the demonstration is defined in a way to generate as much as possible conflicting goals that would challenge/exercise the designed/implemented ADAM SW.
  3. Performing relevant steps ahead with respect the results expected as output of the OG2 to OG6 field test campaign that will take place end of 2018 in Morocco at the scope of pushing forward the current SoA. As first consideration it shall be stressed that the SRC 2016 intends to test the different building block separately one from the other. Thus no integration among OG1 to OG4 is expected at its end. This integration and the required eventual enhancement of OG1 to OG4 towards what this proposal will identify as OG1+ to OG4+ is one of the objectives of this SRC 2018.
  4. Be suitable to easily extend obtained results to terrestrial applications, i.e. deep mining, nuclear, etc.

Demonstration Objectives and System level Considerations

  • The ADE demonstration scenario involves a surface rover equipped with all the software and hardware components compulsory to fulfil the OG10 objectives. In case of ADE the rover avionic will be complemented by an avionics box designed and developed for the scope of OG10. This avionic box hosting the ADAM SW,will use the OG4 sensors and ICU  and required representative on-board processors. ADAM
  • In order to be representative of Mars operations the demonstration scenario encompasses a long traverse (1 km) during 1 sol duration.
  • The impact at system level is the selection of a rover suitable to traverse a minimum of 1 km per day operating for the 6 hours minimum (OG10-R1.1). The robotic platform shall guarantee 15m positioning accuracy per km (OG10-R1.2). The selected surface platform shall be also selected such to be able to move on terrain targeting three different levels of difficulties (low, medium, high) in terms of 1) nature of soil, 2) density of obstacles and 3) slopes. Additionally the rover shall be selected such to present a morphology whose effects, mostly in terms of shadow, shall be such to generate system level impacts affecting the decision making process (OG10-R4.3).
  • Due to this long traverse capability it is evident that the offline planning of the scientific measures becomes almost impossible given the communication windows restriction/delay and the downlink limitations. For these reasons the rover shall be provided with an on-board module capable to autonomously select a scientific interesting target and communicate its position to the autonomous decision module in charge of applying the due re-planning. Within the proposed demonstration scenario the communication windows/rate and the latency typical of a Martian mission shall be taken into account while defining the mission scenario schedule.ADEcomponents

From the point of view of the opportunistic science, the demonstration targets an ambitious scenario where multiple contemporary potential science targets are detected, identified and assessed in terms of their scientific value and collected. It is proposed to foresee within ADAM two science agents based on two independent instruments, a high resolution camera, focused on geometric shape, texture, albedo and specularity, and a thermal camera, with a greater concentration on the thermal signature or discrepancies with respect to the environment. The challenge will be to guarantee high representativeness both at instrument and scientific object detection levels. Introducing multiple sources is a way of increasing de facto the potential science opportunities. At the same time it will allow the ADE team to challenge the ADAM SW with conflicting goals and making it robust against challenging situations. 

  • Last but not least, in order to perform sample collection, the rover will be equipped with a robotic arm that will collect science data (images). The possibility to collect samples via a gripper or sampling device collector will be explored in the frame of the project
  • The scenario shall be demonstrate robustness under uncertainty while exhibiting all typical sources of uncertainty as follow:
    • Presence of obstacles and goals not observable at the initial position.
    • Partial knowledge of the initial observations (e.g. unknown attitude of the lander).
    • Different surface types (sand, rock) with different inclinations.
    • Presence of failures triggered by exogenous events (sand traps, target not reachable, etc.) and internal (resource overconsumption, violation of time limits, hardware failure via simulated injection of errors, etc.).