What is NASA’s CADRE Program and Why is It Important?

A New Era of Exploration

For decades, the story of planetary exploration has been one of solitary pioneers. Robotic explorers like Spirit, Opportunity, Curiosity, and Perseverance—marvels of engineering sent to Mars—were lone emissaries, charting the red dust of an alien world on their own. Each was a self-contained laboratory, a single point of contact between humanity and the unknown. This model of exploration, while enormously successful, has inherent limitations. A single robot, no matter how capable, can only be in one place at a time. Its mission is a single thread of discovery, and its survival represents a single point of failure. A new chapter in robotic exploration is beginning, one founded not on the capabilities of a single, complex machine, but on the emergent power of a team. NASA’s Cooperative Autonomous Distributed Robotic Exploration (CADRE) program is the prologue to this new story.

CADRE is not a flagship science mission in the vein of the grand Mars rovers. It is a technology demonstration, a carefully designed experiment with a singular, focused purpose: to prove that a team of robots can work together on another world, making their own decisions to achieve a common goal. The mission will send a trio of small, wheeled robots to the Moon, but their destination is secondary to their method. The core of the experiment is to give this robotic team a high-level command from Earth—for example, “explore this designated region”—and then stand back as they figure out the rest for themselves. They will need to coordinate their movements, share information, manage their limited resources, and adapt to unexpected challenges as a cohesive unit, all without constant step-by-step instructions from human controllers.

This shift toward autonomy is a direct response to the fundamental constraints of exploring our solar system. While communication with the Moon is nearly instantaneous, with signals taking only a few seconds to travel each way, venturing farther out presents a significant challenge. The light-speed delay to Mars can be as long as 20 minutes each way, making the kind of real-time, joystick-style control used for early lunar rovers impossible. A rover on Mars operating under a “ground-in-the-loop” model, where every decision is vetted by humans on Earth, would spend the vast majority of its time idle, waiting for its next set of commands. This operational bottleneck severely limits the pace and scope of discovery. To explore more distant and dynamic worlds, such as the icy moons of Jupiter or the lava tubes of Mars, robots must possess the intelligence to make decisions on their own. The CADRE mission uses the Moon as a convenient and low-latency proving ground for the very autonomous technologies that are indispensable for the next generation of deep space exploration. Its success will build the confidence needed to deploy similar systems in environments where the communication tether to Earth is stretched to its absolute limit.

Beyond the challenge of communication delays, CADRE also represents a new philosophy of mission design and risk management. A single, highly complex rover is an incredibly valuable asset, and its loss would be a catastrophic end to a mission’s surface operations. A multi-robot system, by its very nature, is more resilient. The CADRE team is designed with the understanding that space exploration is inherently risky. The loss of one rover, while a setback, would not end the mission. The remaining members of the team are programmed to recognize the loss, re-evaluate their plan, and continue their work, still capable of returning valuable data. This concept of “strength in numbers” fundamentally alters the risk calculus for mission planners. It opens the door to more ambitious exploration, allowing a team to send a single member on a high-risk foray into a scientifically compelling but dangerous location, like a steep crater wall or the entrance to a cave. If that scout is lost, the collective mission can still succeed. This distributed approach trades the perfection of a single machine for the resilience of a networked team, paving the way for a more robust and daring era of exploration.

Deconstructing the Mission: The Meaning of CADRE

The name of the mission, CADRE, is an acronym that perfectly encapsulates its core principles. Each word—Cooperative, Autonomous, Distributed, Robotic, Exploration—describes a key facet of this new approach to venturing beyond Earth. Understanding these components reveals the intricate thinking behind the project and its departure from traditional mission architectures.

Cooperative

The first word, “Cooperative,” signifies that these robots are more than just a collection of individuals operating in the same area; they are a true team. Their software is designed to facilitate active collaboration. They don’t simply coexist; they communicate, share data, and coordinate their actions to achieve goals that would be difficult or impossible for any single member to accomplish alone. A primary example of this cooperation is their ability to drive in formation. The rovers will use their onboard radios to constantly measure the distance between one another, allowing the team to maintain a specific geometric pattern as they traverse the lunar landscape. This isn’t just for show; it’s a required capability for certain types of scientific measurements. Another cooperative task is the creation of a unified 3D map of their surroundings. Each rover builds its own local map using its stereo cameras, but by sharing this information through the base station, the team can stitch these individual perspectives together into a single, more comprehensive model of the terrain, filling in gaps and improving overall accuracy. This shared awareness is the foundation of their ability to work together effectively.

Autonomous

“Autonomous” points to the mission’s most significant technological leap: the ability to operate without constant human oversight. For past rover missions, mission controllers on Earth have acted as micromanagers, meticulously planning every movement and action. The CADRE rovers, in contrast, are designed to be self-sufficient problem-solvers. Mission control will issue simple, high-level commands, such as instructing the team to survey a particular patch of ground. From that point on, the rovers take over. Their onboard software will autonomously plan the best routes, navigate around obstacles like rocks and craters, manage their own power and thermal levels, and decide how to divide the labor amongst themselves to complete the task most efficiently. This level of independence dramatically increases the potential science return of a mission by minimizing the time rovers spend waiting for instructions from Earth. It allows the machines to react to their environment in real time, adapting to unforeseen circumstances without waiting for a command to travel hundreds of thousands of kilometers through space.

Distributed

The term “Distributed” refers to the physical separation of the robots and the unique capabilities this separation enables. The CADRE mission is built around the concept of distributed sensing—taking synchronized measurements from multiple vantage points simultaneously. This is the key to unlocking new types of scientific investigation. The mission’s premier example of this is its use of multi-static ground-penetrating radar. A single rover can use radar to probe the subsurface, but the data it gathers is limited. By working as a team, one CADRE rover can transmit a radar signal into the ground while the others listen for the reflections. By precisely measuring the timing and characteristics of these reflected signals from their different positions, the team can build a detailed, three-dimensional map of the lunar subsurface, revealing layers of regolith and bedrock down to a depth of about 10 meters. This kind of tomographic imaging is physically impossible for a single robot to perform. The distributed nature of the system transforms a collection of individual sensors into a single, more powerful, and multi-dimensional instrument.

Robotic Exploration

Finally, “Robotic Exploration” grounds the entire project in its ultimate purpose. At its core, CADRE is a mission of discovery. It uses advanced machines to venture into an alien environment to gather information and expand human knowledge. While the primary goal is to demonstrate the technology of cooperative autonomy, the rovers are still explorers. They will be the first machines to closely survey their landing site at Reiner Gamma, creating detailed surface and subsurface maps. Their journey across the regolith, their negotiation of obstacles, and their systematic data collection embody the fundamental spirit of exploration. The success of the CADRE technology will not be an end in itself; it will be a means to enable more ambitious and far-reaching robotic exploration of the Moon, Mars, and the myriad other worlds in our solar system.

The Robotic Team: Hardware and Instruments

The CADRE mission is composed of a tightly integrated system of hardware, with the three rovers acting as the mobile heart of the operation. These machines, while small, are densely packed with the sensors, computers, and instruments necessary to survive and perform their complex tasks on the lunar surface. Complementing the rovers are the stationary components on the lander that provide deployment, communication, and power infrastructure.

Design and Specifications

At a glance, a CADRE rover is unassuming. Each of the four-wheeled robots is roughly the size of a carry-on suitcase, measuring approximately 0.75 by 0.5 by 0.2 meters with its solar panels deployed, and weighing less than 10 kilograms. This compact and lightweight design is a deliberate choice, driven by the constraints of cost and payload mass on commercial lunar landers. The mobility system is a simple but robust four-wheeled, skid-steer design, where turning is accomplished by varying the speed of the wheels on either side, much like a tracked vehicle. This eliminates the complexity of individual steering motors for each wheel.

Power is the lifeblood of any robotic mission. The CADRE rovers are entirely solar-powered. A top-mounted solar panel, which remains folded during transit, is deployed once the rover is on the surface. This panel charges an internal battery, allowing the rover to perform its tasks. A key design element is the large, upward-facing radiator that also sits on top of the rover’s main chassis. In the airless vacuum of the Moon, radiation is the only effective way to shed waste heat. This radiator is essential for dissipating the heat generated by the rover’s electronics and absorbed from the intense sunlight, preventing the sensitive components from overheating.

The “brain” of each rover is a ModalAI VOXL compute board, which is built around a Qualcomm Snapdragon 821 processor. This is a powerful, quad-core chip with an integrated graphics processing unit (GPU). Its selection is notable because it is a next-generation version of the processor class used in smartphones and, more famously, in NASA’s Ingenuity Mars Helicopter. This reliance on advanced, commercial-off-the-shelf electronics provides immense computational power in a small, energy-efficient package, enabling the complex calculations required for autonomous navigation and coordination. The flight software, written in C++ and built upon NASA’s F’ (FPrime) framework, is designed to take full advantage of this hardware, running processes on both the high-speed and low-speed cores of the CPU and leveraging the GPU for intensive tasks.

The following table provides a summary of the key specifications for each CADRE rover.

This design philosophy reflects a strategic trade-off. Unlike the large, nuclear-powered Mars rovers, which are built to be singular, exquisitely capable, and long-lasting assets, the CADRE rovers are designed with a different purpose in mind. Their hardware is “good enough” for a short, focused technology demonstration. By utilizing lightweight materials, simpler mobility systems, and commercial electronics, the cost and complexity of each individual unit are kept low. This makes it feasible to build and launch a team. The mission’s true power and innovation do not reside in the perfection of any single piece of hardware, but in the sophisticated software that allows these relatively simple machines to act as an intelligent collective. This approach offers a more affordable and scalable model for future planetary exploration.

Sensory and Scientific Payload

Each CADRE rover is equipped with a suite of sensors that allow it to perceive its environment, navigate through it, and perform scientific measurements. These instruments are the eyes, ears, and inner-ear balance of the robotic explorers.

For navigation and mapping, the primary sensors are two pairs of stereo cameras, one facing forward and one facing rearward. Each stereo camera works like a pair of human eyes, capturing two slightly offset images of the same scene. The onboard computer can process these image pairs to determine the distance to various points in the terrain, allowing it to build a 3D map of its surroundings. This map is essential for identifying obstacles like rocks and steep slopes and for planning safe driving paths.

To keep track of its own position and orientation, each rover relies on a suite of navigation sensors. An Inertial Measurement Unit (IMU) measures the rover’s acceleration and rotation, helping it to estimate how far it has moved and in which direction. A sun sensor acts as a compass, determining the rover’s orientation relative to the Sun’s position in the sky. Augmenting these is an ultra-wideband ranging radio. This device sends out signals that are received by the other rovers and the base station. By measuring the time it takes for these signals to travel, the system can calculate the precise distance between agents, which is a key input for maintaining driving formations and for fusing together their individual maps.

The mission’s key scientific instrument is the multi-static Ground-Penetrating Radar (GPR). The GPR antenna is mounted on the front underside of each rover, facing the ground. In a “monostatic” mode, a single rover can send a radio pulse into the regolith and listen for the echo, which provides a one-dimensional profile of the subsurface directly beneath it. The true innovation of CADRE is its “multistatic” mode. In this configuration, one rover acts as a transmitter, sending the radar pulse, while the other two rovers act as receivers, listening for the signals that reflect off of subsurface layers. Because the receivers are at different, known distances from the transmitter, the system can combine their data to create a high-resolution, three-dimensional image of the subsurface geology. This cooperative technique allows the team to probe for features like buried rock layers or potential ice deposits with a level of detail that a single rover could never achieve.

The Full Ensemble

The three rovers are the stars of the show, but they are part of a larger system. A crucial component is the base station, which remains fixed to the deck of the lunar lander. The base station is equipped with the same type of computer and mesh network radio as the rovers but has no wheels or science instruments. It serves two main functions. It acts as the primary communication gateway, relaying data from the rover network back to Earth and transmitting commands from Earth to the rovers. It also serves as a central computing hub. The rovers can offload computationally intensive tasks, like merging their individual maps into a single global map, to the base station, which has a more stable power supply from the lander.

Getting the rovers from the lander’s deck to the lunar surface is the job of the deployer systems. There is one deployer for each rover. The system consists of a motorized spool that holds a strong, thin fiber tether. After landing, a release mechanism frees the rover, and the spool slowly and carefully unwinds the tether, lowering the rover to the ground. The deployment is designed to be deliberately slow, allowing ground control to monitor the process and ensure a safe touchdown. Once on the surface, the rover detaches from the tether and drives away.

Finally, a situational awareness camera assembly, also mounted on the lander, will act as an impartial observer. This camera will monitor the rovers as they deploy and conduct their experiments on the surface, providing a visual record of their activities and giving engineers on Earth a valuable overview of the mission’s progress. Together, these components—rovers, base station, deployers, and camera—form a complete, self-contained system for demonstrating cooperative autonomous exploration.

The Brains of the Operation: A New Generation of Autonomy

The hardware of the CADRE mission provides the physical means for exploration, but the true innovation lies in its software. The ability of three separate robots to think and act as a single, coordinated entity is the result of a sophisticated, multi-layered autonomy architecture. This system is designed to solve the immense challenge of managing a team of robots in a remote, unpredictable, and resource-constrained environment like the Moon. It must handle everything from high-level strategic planning to the low-level execution of a single wheel rotation, all while ensuring the team works together harmoniously and resiliently.

The Challenge of Teamwork in Space

Coordinating a robotic team on another world presents a unique set of problems. The lunar surface is an unstructured environment, filled with hazards like craters, rocks, and soft regolith that can trap a rover. Communication between the rovers can be intermittent, blocked by terrain features or affected by the properties of the lunar dust. Furthermore, each rover has its own finite and constantly changing resources to manage, primarily its battery charge and internal temperature. A plan that works for a fully charged rover in the cool lunar morning might be impossible for a rover with a low battery in the heat of midday.

A successful autonomy system must be able to create a plan that respects all these constraints, not just for one robot, but for the entire team. It needs to ensure that tasks requiring cooperation, like driving in formation, are synchronized. It must also be robust against failure. What happens if one rover’s communication link drops out? What if a rover gets stuck or its battery dies unexpectedly? The software architecture must be able to detect these problems, adapt the team’s plan, and continue the mission with the remaining assets. The CADRE system is NASA’s answer to these complex challenges.

The CADRE Software Architecture

The core of the mission’s intelligence is its Planning, Scheduling, and Execution (PS&E) system. This software is built upon a philosophy of “centralized planning, distributed execution.” This hybrid approach is a pragmatic solution to the challenges of the lunar environment. Instead of a fully decentralized system where every rover is constantly planning and negotiating with its peers—a process that would require immense communication bandwidth—the CADRE team centralizes the difficult task of long-term planning onto a single agent at any given time. the execution of that plan is distributed, with each rover responsible for carrying out its own assigned tasks. This reduces the communication overhead while still allowing for individual flexibility.

The architecture is a direct and elegant solution to the constraints of the operating environment. A purely decentralized system would fail with the Moon’s unreliable communication links. A purely centralized system with a fixed leader would fail if that single leader were to be lost. The CADRE architecture finds a middle ground. It gains the efficiency of centralized planning by designating a leader, but it builds in resilience by making that leadership role dynamic and transferable. The leader election module and the designated survivor concept act as the system’s self-healing mechanism, allowing the team to recover from the loss of its current “brain” and seamlessly continue the mission. This design trades the need for constant, perfect communication for a more robust and fault-tolerant system.

Electing a Leader

The CADRE team does not have a permanent, fixed leader. Instead, it employs a dynamic leader election process to designate one agent—which can be one of the three rovers or the base station—to act as the team’s planner. This is not a vote based on popularity; it is a selection based on fitness. The role of the leader is computationally intensive and consumes more power and generates more heat than other tasks. Therefore, the system continuously assesses which agent is in the best condition to take on this demanding role.

Every ten seconds, the agents engage in a distributed algorithm to elect a unique “appointer.” This appointer gathers health and status data from all other members of the team, looking at factors like battery state-of-charge and CPU temperature. It then selects the agent that is furthest from violating any of its operational limits—in other words, the healthiest robot—to be the new leader. This designation is then broadcast to the entire team. To prevent the leadership from changing too frequently, a hysteresis mechanism is built in, meaning the leader will not be replaced unless there is a significant change in its status.

To further enhance robustness, the system also selects a “designated survivor.” This agent is tasked with maintaining a backup copy of the leader’s state and the current team plan. If the leader unexpectedly fails or loses communication, the team can quickly promote a new leader, which can then retrieve the latest plan from the designated survivor. This minimizes downtime, ensuring that the loss of the leader results in at most a ten-second pause before a new leader takes over and planning resumes.

Sharing Knowledge

For the elected leader to create an effective plan for the team, it needs to have an up-to-date picture of what the entire team is doing and what condition it’s in. This is the function of the shared state database, a system called MoonDB. It is a lightweight, distributed database designed specifically for the CADRE mission.

Each agent maintains its own local database of information, including its position, its power and thermal status, and the local terrain map it has generated. The MoonDB system is responsible for synchronizing a select subset of this information from each rover to the elected leader. It’s designed to work over the intermittent and limited-bandwidth communication links on the Moon. It doesn’t attempt to keep every piece of data on every rover perfectly synchronized at all times. Instead, it uses judicious policies to decide what information is most important to share and how often to share it. For example, critical data like a rover’s position might be updated more frequently than less time-sensitive information.

MoonDB also performs data fusion. When the leader receives local maps from each of the other rovers, MoonDB helps to merge them into a single, consistent global map of the explored area. This shared world model is what allows the leader to plan coordinated movements and assign tasks that make sense for the team as a whole. It provides the leader with a unified, coherent view of the team and its environment, which is the necessary foundation for intelligent multi-agent planning.

Executing the Plan

Once the leader has been elected and has received the latest state information from the team via MoonDB, the strategic planning module gets to work. This module, which uses a planning and scheduling tool called MEXEC, is the core decision-making engine. It takes the high-level goal from Earth (e.g., “perform a GPR survey of this area”), considers the current state of all rovers (their location, resources), and generates a detailed, time-ordered sequence of tasks for each agent. This plan might include tasks like “Rover 1, drive to waypoint X,” “Rover 2, hold position,” and “All rovers, activate GPR at time Y.” The planner’s key function is to create a schedule that accomplishes the mission’s objectives while respecting all known constraints, such as ensuring no rover’s battery will be depleted or its CPU will overheat.

After the strategic planner finalizes the team-wide plan, it is distributed to each individual rover. On each rover, a local agent controller module receives its assigned portion of the plan. This controller is responsible for the final step: translating the abstract command into actions for the rover’s hardware. For example, it takes the command “drive to waypoint X” and passes it to the rover’s single-agent mobility and navigation software, which then calculates the necessary wheel movements to get there while avoiding any immediate, unforeseen obstacles. The agent controller continuously monitors the rover’s progress in executing its tasks and sends status updates back to the leader. If a task is completed successfully, or if it fails for some reason, that information is reported back, allowing the leader to update its model and, if necessary, generate a new plan for the team. This closed loop of planning, execution, and feedback is what allows the CADRE team to operate autonomously and adaptively on the lunar surface.

The Mission to Reiner Gamma

The CADRE technology demonstration is not just a software experiment; it is a fully-fledged space mission with a specific destination and a detailed plan of operations. The journey from the clean rooms of the Jet Propulsion Laboratory to the dusty plains of the Moon involves a complex interplay of commercial partnerships, cutting-edge launch technology, and a carefully chosen target that is both scientifically intriguing and an ideal stage for the rovers’ debut.

Journey to the Moon

The CADRE rovers and their associated hardware will not be traveling to the Moon on a dedicated NASA spacecraft. Instead, they will be flying as a payload on a commercial mission. The launch will take place from Florida’s Kennedy Space Center aboard a SpaceX Falcon 9 rocket. This rocket will carry the Intuitive Machines Nova-C lander, a privately developed spacecraft, on its third mission to the Moon, designated IM-3. The CADRE hardware will be integrated onto the deck of this lander for the journey through space.

This arrangement is made possible by NASA’s Commercial Lunar Payload Services (CLPS) initiative. Through CLPS, NASA acts as a customer, purchasing space for its science instruments and technology demonstrations on missions operated by American companies. This approach is intended to stimulate the development of a commercial lunar transportation market, reducing costs for the agency while creating a regular cadence of missions to the Moon. The IM-3 mission, carrying CADRE and other NASA payloads, is a prime example of this public-private partnership model in action, forming a key part of the logistical backbone for the broader Artemis program. The launch is currently slated for a window extending into early 2026.

A Mysterious Destination: The Lunar Swirls of Reiner Gamma

The destination for the IM-3 lander and its robotic passengers is Reiner Gamma, a feature located in the vast, dark lava plains of Oceanus Procellarum (the Ocean of Storms) on the Moon’s nearside. Reiner Gamma is one of the most prominent and enigmatic of the Moon’s “lunar swirls.” Visible even through backyard telescopes, it appears as a bright, sinuous, almost painted-on pattern that stands in stark contrast to the surrounding dark mare.

For years, the origin of these swirls was a mystery. The leading scientific hypothesis now links them to the presence of localized magnetic anomalies—pockets of ancient magnetic fields embedded in the lunar crust. The Moon as a whole does not have a global magnetic field like Earth’s, but these remnant fields are strong enough to interact with the solar wind, the constant stream of charged particles flowing from the Sun. This magnetic field acts like a miniature, invisible shield or “sunscreen.” It deflects the incoming solar wind particles, preventing them from bombarding the surface directly beneath the anomaly. Over billions of years, the solar wind darkens the lunar soil, a process known as space weathering. The areas protected by the magnetic field remain relatively pristine and bright, while the surrounding regions darken, creating the distinctive swirl patterns.

This unique environment makes Reiner Gamma a natural laboratory for studying the complex interactions between planetary surfaces, magnetic fields, and space plasma. It offers a chance to investigate the history of the Moon’s magnetic field, the processes of space weathering, and the very nature of these strange geological features.

The selection of Reiner Gamma for this mission is a case of strategic synergy. While CADRE’s primary goal is to test autonomy, the IM-3 lander will also be carrying another key NASA payload: the Lunar Vertex mission. Lunar Vertex is a dedicated science mission, consisting of a lander-based magnetometer and a small rover, designed specifically to study the magnetic fields and plasma environment of Reiner Gamma to solve the mystery of the swirls. By sending CADRE to the same location, NASA creates a powerful, complementary scientific opportunity. While Lunar Vertex measures the fields and particles above and on the surface, the CADRE team’s ground-penetrating radar will be mapping the subsurface geology. The structure of the crust beneath the swirl is a key piece of evidence needed to understand the origin of the magnetic source itself. Together, these two missions will provide a far more complete picture of this fascinating lunar feature than either could achieve alone, demonstrating an efficient and integrated approach to science under the CLPS program.

A Lunar Day of Experiments

The CADRE mission on the surface is designed to last for the duration of one lunar day, which is equivalent to about 14 Earth days. The lander is scheduled to touch down shortly after local sunrise, and the mission will conclude when the long, cold lunar night begins.

The surface operations will begin with the deployment sequence. One by one, the three rovers will be gently lowered from the lander’s deck to the lunar regolith on their individual tethers. Once a rover’s wheels touch down, it will detach from its tether, perform a system checkout, and drive a short distance away from the lander to a pre-planned “sunbathing” spot. Here, it will deploy its solar panel and begin charging its batteries in preparation for the main experiments.

Once all three rovers are charged and ready, the team will embark on a series of carefully planned experiments designed to validate their cooperative autonomous capabilities. The first test will be to drive in formation. The rovers will use their ultra-wideband radios to maintain a precise distance from one another as they traverse the terrain, demonstrating their ability to coordinate their movements as a single unit.

The next major experiment will be a cooperative mapping task. The team will be given a command to explore a designated area of about 400 square meters (roughly 4,300 square feet). The rovers will then autonomously devise their own paths to survey the region, using their stereo cameras to build a combined 3D topographic map. During this phase, the project will also test the team’s resilience by simulating the failure of one rover. The software on the remaining two rovers is expected to recognize that their teammate is no longer functioning, adapt their exploration plan, and complete the mapping task on their own.

The final and most scientifically significant experiment will be the multi-static GPR survey. Driving in their coordinated formation, the rovers will systematically scan the area, using their combined radar capabilities to build a 3D map of the subsurface structure. This data will not only demonstrate the power of distributed sensing but will also provide valuable geological context for the Reiner Gamma swirl, contributing directly to the scientific goals of the overall IM-3 mission. Throughout all these activities, the camera on the lander will be watching, providing a visual record of this pioneering experiment in robotic teamwork.

Surviving the Moon: Overcoming Environmental Hurdles

The lunar surface is one of the most hostile environments in the solar system, and any machine designed to operate there must be engineered to withstand a host of extreme challenges. For the small, lightweight CADRE rovers, two environmental factors dominate the design and operational planning: the brutal cycle of temperature extremes and the pervasive, insidious nature of lunar dust.

The Tyranny of Temperature

Operating near the Moon’s equator, the CADRE rovers will face a thermal environment of incredible contrasts. With no atmosphere to moderate temperatures, the surface can swing from frigid cold during the night to scorching heat at midday. During their 14-day operational period, the rovers will have to endure direct sunlight that can heat the surface to over 114°C (237°F).

This external heat load is only part of the problem. The rovers’ own electronics, especially the powerful Snapdragon processor that serves as their brain, generate a significant amount of waste heat during operation. In a vacuum, this heat cannot be carried away by convection as it would be on Earth. It must be radiated away into space. This is the job of the rover’s top-mounted radiator. for a robot as small and compact as a CADRE rover, there is a limit to how much heat can be shed. The combination of intense solar heating and internal heat generation poses a serious risk of overheating the sensitive electronics.

To solve this problem, the CADRE team devised an ingenious operational strategy: a series of autonomous, synchronized “wake-sleep” cycles. Instead of attempting to operate continuously through the hottest part of the lunar day, the rovers are programmed to work in 30-minute intervals. They will power up, perform their exploration and mapping tasks for about half an hour, and then, as a team, they will all shut down. During this “sleep” period, they cease most operations, allowing their radiators to cool their internal components and their solar panels to recharge their batteries. After another 30 minutes, they will simultaneously awaken, re-establish communication, and continue their mission.

This wake-sleep cycle is not merely a simple thermal precaution; it is a complex autonomy challenge that the software must manage at a team level. For cooperative tasks to work, all rovers must be active at the same time. One rover sleeping while others are trying to drive in formation would break the experiment. The PS&E software, and specifically the Strategic Planner on the leader rover, must therefore incorporate these thermal limits into its team-wide schedule. It has to plan activities around these mandatory, synchronized sleep periods. This transforms a fundamental hardware limitation into a sophisticated, software-driven resource management problem, providing a perfect real-world test of the autonomy system’s ability to juggle multiple, conflicting constraints across the entire team.

The Problem of Dust

Lunar dust, or regolith, is unlike any dust found on Earth. It is not the soft, rounded particulate matter of our world. It is the product of billions of years of micrometeorite impacts shattering lunar rock in an airless environment. The resulting particles are small, sharp, and highly abrasive, with a consistency more like crushed glass than sand. Furthermore, due to the constant bombardment of solar radiation and the lack of atmospheric moisture, these particles are electrostatically charged, causing them to cling tenaciously to any surface they touch.

The Apollo astronauts discovered firsthand how problematic this dust could be. It clogged the joints of their spacesuits, abraded visors, caused thermal control surfaces to overheat, and was inadvertently tracked into the lunar module, causing irritation and mechanical issues. For a long-duration robotic mission like CADRE, lunar dust presents several serious threats. Its abrasive nature can wear down moving parts like wheel mechanisms. Its tendency to cling can coat solar panels, reducing their power output, and cover radiators, degrading their ability to shed heat by altering their carefully engineered optical properties. Fine particles can also work their way into seals and other sensitive components, causing them to fail.

The CADRE rovers have been designed with dust tolerance in mind. Their spoked wheels, similar in design to those on NASA’s larger VIPER rover, are intended to minimize the amount of dust kicked up during driving. Sensitive mechanisms are sealed or protected where possible. there is no perfect solution to the dust problem. The mission’s short duration of 14 days helps to mitigate the worst of the long-term effects, but the performance of the rovers in this dusty environment will be closely monitored and will provide valuable data for the design of future, longer-lasting lunar surface missions.

From Concept to Reality: Testing and Validation

The journey of the CADRE project from an idea on a whiteboard to a set of flight-ready robots prepared to land on the Moon has been a multi-year process of design, prototyping, and rigorous testing. Every piece of hardware and every line of code has been subjected to a comprehensive verification and validation campaign to ensure the system is ready for the unforgiving environment of space and the complexities of its autonomous mission.

A Legacy of Innovation: From A-PUFFER to CADRE

The core concepts behind CADRE did not emerge from a vacuum. They are the direct descendants of a previous research and development project at JPL known as A-PUFFER, the Autonomous Pop-Up Flat Folding Explorer Robot. The A-PUFFER project developed small, two-wheeled robots designed to fold nearly flat, allowing them to be deployed in hard-to-reach places. More importantly, A-PUFFER served as the initial testbed for the multi-agent autonomy software that forms the foundation of CADRE’s intelligence. This technological heritage demonstrates that CADRE is not a radical shot in the dark, but rather the culmination of a deliberate, evolutionary process of advancing the state of the art in cooperative robotics for space exploration. The lessons learned from A-PUFFER in software architecture, communication protocols, and autonomous behaviors were directly infused into the CADRE project, giving it a running start on its path to the Moon.

Forging a Lunar Team on Earth

Long before the actual flight rovers were built, their software and hardware designs were put through their paces in simulated lunar environments on Earth. A key facility for this testing is JPL’s Mars Yard, an outdoor test area whose rocky, sandy terrain is designed to mimic the surface of other worlds. Here, engineers used full-scale “development models” of the CADRE rovers to test the system’s capabilities in a realistic setting.

These development models, while not built with space-grade components, are identical in size, shape, and mobility to the flight rovers and run the same autonomy software. In the Mars Yard, the team successfully demonstrated the rovers’ ability to perform their key mission tasks. They drove in formation across uneven ground, maintaining their relative positions. When presented with unexpected obstacles, they cooperatively shared updated map data and replanned their paths as a group to find a safe way forward. The tests also validated their adaptive behaviors. In one scenario, engineers artificially drained the battery of one rover. The team’s software correctly identified the low-power state of its companion, and the entire group paused its exploration, waiting until the rover’s battery could be notionally recharged before continuing the mission together. These tests, some conducted at night under powerful floodlights to simulate the harsh shadows of the lunar daytime, were essential for proving that the complex hardware and software systems could work together to achieve the mission’s goals.

Surviving the Journey

While the Mars Yard tests validated the rovers’ performance, a separate and equally important series of tests was needed to ensure the flight hardware could survive the trip to the Moon. This “shake and bake” qualification process subjects the completed rovers to the extreme conditions they will experience during launch and in space.

To simulate the violent vibrations and g-forces of a rocket launch, each rover was clamped to a specialized “shaker table.” This machine vigorously shook the rover in multiple axes, ensuring that all its components were securely mounted and that no electrical connections would come loose during the jarring ride out of Earth’s atmosphere.

To simulate the environment of space and the lunar surface, the rovers were placed inside a thermal vacuum (TVAC) chamber. In this chamber, all the air is pumped out to create a near-perfect vacuum. The temperature inside is then cycled between extreme highs and lows, mimicking the transition from direct sunlight to deep shadow. This testing confirms that the rovers’ thermal control systems, including their radiators and heaters, function as designed and that all electronic components can operate reliably across the full range of expected temperatures. The rovers also underwent electromagnetic compatibility testing to ensure that their own electronic subsystems did not interfere with each other or with the sensitive electronics of the lander.

The “Mercury 7” Prototypes

Even before the full-scale development models were built, the CADRE team used a set of smaller, plastic prototypes for early-stage software and algorithm development. These seven test rovers were affectionately nicknamed the “Mercury 7” by the team, with each one named after one of NASA’s original Project Mercury astronauts. These prototypes were used in sandboxes and other lab environments to test fundamental capabilities like the formation-driving algorithm. This nod to the pioneers of American human spaceflight underscores the project’s deep connection to the history of exploration and its role in paving the way for the next generation of explorers, both robotic and human.

The Multi-Robot Advantage

The decision to send a team of three small rovers to the Moon instead of one larger, more capable rover was a deliberate strategic choice. The multi-robot approach offers a suite of distinct advantages over the traditional single-explorer model, enabling new kinds of science and providing a more resilient and efficient platform for exploration. These benefits are not just theoretical; they are the very reason the CADRE mission was conceived.

Beyond the Lone Explorer

A single rover, like NASA’s Perseverance on Mars, is a masterpiece of engineering, capable of conducting a wide range of sophisticated scientific analyses. its exploration is inherently linear. It moves from one point of interest to the next, collecting data along a single path. To map a large area, it must painstakingly traverse back and forth in a “lawnmower” pattern, a time-consuming process. Its scientific measurements are limited to what can be done at a single point in space and time. The CADRE team, by contrast, operates in parallel, fundamentally changing the nature of robotic exploration.

Strength in Numbers

The benefits of the multi-robot approach can be broken down into four key areas: efficiency, fault tolerance, novel science capabilities, and scalability.

First, a team of robots is far more efficient at mapping tasks. Three rovers exploring a designated area can cover the ground roughly three times faster than a single rover. They can divide the area into sectors, with each member responsible for mapping its own zone simultaneously. This parallel approach dramatically reduces the time required to create a comprehensive survey of a landing site, a capability that will be invaluable for future missions where rapid reconnaissance is needed to support astronaut activities or identify resource deposits.

Second, the team possesses inherent fault tolerance and resilience. In the high-risk endeavor of space exploration, the possibility of failure is always present. A single rover getting stuck in soft regolith or suffering a critical hardware failure could mean the premature end of a billion-dollar mission. For the CADRE team, the loss of a single agent is not a mission-ending event. The remaining rovers are programmed to adapt to the loss of a teammate and continue the mission, albeit with reduced capability. This built-in redundancy makes the overall mission more robust and allows for the consideration of higher-risk exploration strategies that would be unthinkable with a single, irreplaceable asset.

Third, a distributed team can perform novel science measurements that are physically impossible for a single robot. The premier example for CADRE is multi-static ground-penetrating radar, which requires a separated transmitter and multiple receivers to build a 3D subsurface image. This principle extends to other potential scientific investigations. A network of rovers could deploy seismometers over a wide area to create a lunar seismic network, allowing scientists to triangulate the source of moonquakes and study the Moon’s deep interior. A team of rovers equipped with telescopes could be arranged to form a synthetic aperture, combining their light-gathering power to function as a single, much larger telescope. These distributed sensing techniques open up entirely new avenues for planetary science.

Finally, the CADRE architecture is designed for scalability. While this initial technology demonstration is limited to three rovers due to the mass and budget constraints of the mission, the underlying software and coordination principles are not. The leader-election and task-planning algorithms can be scaled to accommodate a much larger number of agents. Future missions could deploy dozens of rovers to conduct large-scale resource prospecting campaigns or to work as a construction crew, building landing pads and habitats from lunar regolith. CADRE is laying the groundwork for these future “swarm” missions, proving that the foundational technology for coordinating a large robotic workforce is viable.

Paving the Way for the Future: CADRE’s Role in NASA’s Vision

The CADRE mission, while a self-contained technology demonstration, does not exist in a vacuum. It is a vital and intentionally designed component of NASA’s broader, long-term strategy for the exploration of the solar system. Its success will have direct implications for the Artemis program and the agency’s overarching Moon to Mars architecture, serving as a key enabling technology for a future of sustained human and robotic presence beyond Earth.

Supporting the Artemis Generation

NASA’s Artemis program is the agency’s ambitious plan to return humans to the Moon, this time to establish a sustainable, long-term presence. This vision involves not just short sorties to the surface, but the construction of an Artemis Base Camp, the use of lunar resources, and a continuous cadence of scientific research and technological development. In this vision, robots are not just precursors; they are essential partners to human astronauts.

Technologies like those being proven by CADRE are foundational to this future. Teams of autonomous robots will be able to perform many of the dull, dirty, and dangerous tasks required to build and maintain a lunar outpost. They can conduct extensive geological surveys to identify the best locations for habitats or to prospect for resources like water ice. They can transport equipment, perform routine inspections of infrastructure, and even assist in construction activities. By offloading these tasks to autonomous robotic teams, the limited and highly valuable time of astronauts on the surface can be reserved for complex scientific investigations and other activities that require human ingenuity and dexterity. CADRE’s delivery to the Moon as part of the CLPS program places it directly within the support infrastructure of the Artemis campaign. It is a test flight for the kind of robotic assistants that will one day work alongside Artemis astronauts.

The success of CADRE directly supports several core tenets of NASA’s official Moon to Mars Objectives. The agency’s strategy emphasizes the development of integrated human and robotic methods, the creation of a sustainable lunar infrastructure, and the maximization of crew time available for science. An autonomous robotic workforce, whose foundational operating system is being tested by CADRE, is a direct answer to these objectives. Future, scaled-up versions of this technology are critical for the “Live” and “Explore” elements of NASA’s Lunar Surface Innovation Initiative, which includes the large-scale excavation, construction, and resource utilization tasks that are perfectly suited for autonomous robot teams. CADRE is not merely a passenger on an Artemis-era flight; it is a validation of a key architectural component of the entire Moon to Mars plan.

A Proving Ground for Mars and Beyond

The ultimate goal of NASA’s human exploration program is to send the first astronauts to Mars. The Moon serves as a crucial proving ground for this monumental undertaking, a place to test the technologies and operational concepts needed for the much longer and more challenging journey to the Red Planet. This is especially true for autonomous systems.

The significant communication delay between Earth and Mars makes the kind of real-time human oversight possible for lunar missions completely impractical. A robotic team on Mars must be able to operate on its own for hours or even days at a time. Proving that the CADRE system of cooperative autonomy works reliably in the relatively close and familiar environment of the Moon is a necessary prerequisite before entrusting a similar system to a multi-year mission to Mars, where there is no possibility of quick intervention or rescue.

The technology demonstrated by CADRE is also highly portable. The core software architecture for multi-agent coordination is not specific to the Moon or to rovers. The same principles could be applied to a team of autonomous drones flying in the thin Martian atmosphere, a network of cryobots exploring the subsurface ocean of Europa, or a mixed team of rovers and drones working together. By validating the fundamental concepts on the Moon, CADRE is developing a versatile exploration toolkit that can be adapted for a wide variety of future missions across the solar system.

The Future of Autonomous Exploration

CADRE is the first step, not the final word, in cooperative robotic exploration. Its success will open the door to a future where swarms of intelligent machines are deployed to tackle tasks of increasing scale and complexity. One can envision future missions where dozens of CADRE-like rovers are dispatched to the lunar poles to conduct a comprehensive prospecting campaign, systematically mapping out the location and concentration of water ice deposits—a resource essential for life support and rocket fuel.

Further into the future, teams of larger, more powerful robots based on this technology could become the primary construction crews for a lunar base. They could autonomously excavate foundations, 3D-print structures from lunar regolith, assemble habitats, and lay down power and communication lines, all before the first human inhabitants arrive. When astronauts are present, these robotic teams could function as assistants, transporting tools, collecting samples from designated locations, and performing maintenance, all in response to high-level voice commands.

This vision of the future, where humans and robots work as true partners in the exploration and settlement of space, begins with the three small rovers of the CADRE mission. They are not just mapping a small patch of the Moon; they are charting a course toward a new way of exploring the cosmos, one defined by teamwork, shared intelligence, and the boundless potential of a robotic wolf pack.

Summary

NASA’s Cooperative Autonomous Distributed Robotic Exploration mission represents a fundamental shift in the methodology of exploring other worlds. Moving beyond the paradigm of the solitary rover, CADRE is a technology demonstration designed to prove the viability and effectiveness of a team of robots working together without direct, continuous human command. The mission’s name encapsulates its core tenets: a Cooperative team that works together, an Autonomous system that makes its own decisions, a Distributed network that enables novel measurements, and a focus on Robotic Exploration of the lunar surface and subsurface.

The mission will deploy a trio of suitcase-sized, solar-powered rovers to the enigmatic Reiner Gamma region of the Moon. There, over the course of a single lunar day, they will execute a series of experiments to validate their groundbreaking software. This software architecture, based on a “centralized-planning, distributed-execution” model, features a dynamic leader election process that ensures both efficiency and resilience. The rovers will drive in formation, cooperatively map the terrain, and use their combined ground-penetrating radars to create a 3D image of the subsurface—a feat impossible for a single robot.

CADRE is engineered to overcome the harsh lunar environment, employing an innovative, autonomous wake-sleep cycle to manage extreme temperatures and a robust design to tolerate the abrasive lunar dust. The system has undergone extensive testing on Earth, from early prototypes to full-scale models in simulated lunar landscapes, to ensure its readiness for the challenges ahead.

The success of this mission will provide more than just a proof of concept. It will validate a technology that is a key enabler for NASA’s long-term vision, directly supporting the goals of the Artemis program by developing the robotic assistants that can help establish a sustainable human presence on the Moon. It serves as an essential proving ground for the autonomy required for future, more complex missions to Mars and beyond. CADRE is the first chapter in a new story of exploration, one where the answer to the question “What can we discover?” is determined not by a single explorer, but by what a team can accomplish together.