Planning, scheduling, and execution on the Moon: the CADRE technology demonstration mission

TL;DR

CADRE tackles the challenge of coordinating four autonomous agents for lunar surface exploration and subsurface sensing under limited communication and energy constraints. It introduces a leader-election-based PS&E stack with a lightweight SSDB, centralized planning at the leader, and decentralized execution using multi-agent constraint checking, with planning at $1\mathrm{Hz}$ and leader elections every $10\mathrm{s}$, within $25$-minute wake cycles and CPU temperature limits around $65^\\\\\\circ$C. The architecture minimizes inter-agent messaging while preserving safety through continuous constraint checks and an abort mechanism, and is validated through ROS simulations, Dragonfarm, and flight-model testing, demonstrating end-to-end feasibility for Moon deployment. Overall, the work advances technology readiness for multi-agent planetary exploration and offers a scalable framework for future science-driven lunar missions.

Abstract

NASA's Cooperative Autonomous Distributed Robotic Exploration (CADRE) mission, slated for flight to the Moon's Reiner Gamma region in 2025/2026, is designed to demonstrate multi-agent autonomous exploration of the Lunar surface and sub-surface. A team of three robots and a base station will autonomously explore a region near the lander, collecting the data required for 3D reconstruction of the surface with no human input; and then autonomously perform distributed sensing with multi-static ground penetrating radars (GPR), driving in formation while performing coordinated radar soundings to create a map of the subsurface. At the core of CADRE's software architecture is a novel autonomous, distributed planning, scheduling, and execution (PS&E) system. The system coordinates the robots' activities, planning and executing tasks that require multiple robots' participation while ensuring that each individual robot's thermal and power resources stay within prescribed bounds, and respecting ground-prescribed sleep-wake cycles. The system uses a centralized-planning, distributed-execution paradigm, and a leader election mechanism ensures robustness to failures of individual agents. In this paper, we describe the architecture of CADRE's PS&E system; discuss its design rationale; and report on verification and validation (V&V) testing of the system on CADRE's hardware in preparation for deployment on the Moon.

Figures (6)

• Figure 1: Overview of the CADRE mission.
• Figure 2: Architecture of CADRE's flight software. The second and third row of blocks from the top represent the PS&E architecture.
• Figure 3: Exploration: two rovers (magenta dots) explore the sub-region assigned to them (lighter rectangle) by selecting a frontier (colorful segments) between traversable (off-white) and unexplored (light gray) portions of the sub-region.
• Figure 4: Distributed sensing: each rover is assigned a "tube" around its nominal trajectory, and must travel inside the tube while satisfying temporal constraints.
• Figure 5: Abstract representation of the task network for both formation sensing and exploration. Arrows represent precedence constraints. Resource and multi-agent constraints are not shown.