Artemis 2: An Architectural Critique

The recent Artemis 2 mission, which cost approximately $4.1 billion, sent humans to orbit the Moon. A flight profile executed 10 times since the Apollo era. Evaluating this from an engineering perspective, sending a crewed mission without first establishing the necessary communication and navigation infrastructure is a fundamental misallocation of resources.

Architectural Flaw 1: Infrastructure Sequencing

During the Orion spacecraft's pass behind the Moon, the mission experienced a 40-minute communication blackout. The live video feed quality was extremely poor due to the bandwidth limitations of the legacy S-band network. In my first article, "Road To A Lunar Base" (November 6, 2024), I established that a relay satellite network is the absolute prerequisite for lunar operations. A base positioned at the Compton-Belkovich Thorium Anomaly (CBTA) on the far side requires continuous line-of-sight communication. The 40-minute blackout during Artemis 2 proves that the Lunar Communications Relay and Navigation Systems (LCRNS) must be deployed before human missions.

NASA claims the astronauts gathered finer details of the lunar surface using their eyes and handheld Nikon cameras compared to previous robotic missions. This statement lacks technical validity. The Lunar Reconnaissance Orbiter (LRO) currently provides a global resolution of 0.5 meters per pixel. Human vision is limited to the 400–700nm spectrum; a human cannot detect titanium, thorium, or water ice. A dedicated robotic platform equipped with synthetic aperture radar (SAR) and multispectral sensors in a Sun-synchronous polar orbit provides data orders of magnitude superior to visual observation.

The argument that human presence is required for the "gaze factor" or real-time decision-making is outdated. Autonomous edge computing has solved this problem. An AI unit utilizing saliency mapping algorithms can process visual inputs and identify high-albedo anomalies or geometric deviations without human latency. Astronaut reaction time actually increases in microgravity. A radiation-hardened processor operating on a few watts from solar panels reacts in microseconds, directing gimbals and capturing high-resolution data while performing on-device filtering. Instead of transmitting low-quality video, the AI processes terabytes locally and downlinks only the high-value scientific anomalies.

Architectural Flaw 2: Interim Hardware Testing

The mission architecture of Artemis 2 presents another engineering flaw. During the Apollo program, the Saturn V was developed as a complete, final-design solution capable of delivering 48 metric tons to Trans-Lunar Injection (TLI). It was operated as a unified stack. In contrast, the Space Launch System (SLS) used for Artemis operates on an interim Block 1 configuration. While it generates 39MN of liftoff thrust—exceeding the Saturn V—its current payload capacity to TLI is limited to 27 metric tons due to its reliance on a temporary upper stage (ICPS).

Testing an interim configuration with human payloads violates basic engineering efficiency. The logical approach is to finalize the ultimate architecture, specifically the configurations utilizing the Exploration Upper Stage (EUS), and conduct tests exclusively on the definitive hardware. Iterative testing of obsolete rocket configurations wastes resources. A final, integrated design must be established and tested as a complete system before executing crewed missions.

Architectural Flaw 3: Payload Decoupling

The ultimate objective of the Artemis program is to establish a sustained lunar base, which fundamentally differentiates it from Apollo. However, the current mission architecture contradicts this goal. A sustainable engineering roadmap dictates that heavy payloads, construction machinery, and research rovers must be pre-deployed to the target location via autonomous cargo landers prior to human arrival.

This strategy provides two critical technical advantages. First, it drastically reduces the payload requirements for the human-rated vehicles. Life support systems and crew accommodations already consume a massive portion of the mass budget; forcing the crewed vehicle to carry base operations equipment simultaneously is inefficient. Second, descent and landing represent the most critical and failure-prone phases of any mission. Deploying robotic payloads in advance serves as a direct validation of the specific landing zone parameters. The required infrastructure and hardware must be delivered and verified on the surface first, allowing subsequent human missions to arrive at a functional site.