Recent work by Kumar et al. (2025, Adv. Space Res.) has demonstrated the key role of lunar mean motion resonances (MMRs) in shaping the dynamical structure of cislunar space. These MMRs contain unstable orbits with attached stable and unstable manifolds, the intersections of which generate heteroclinic “highways” that spacecraft can follow from much lower orbits all the way to the Moon for zero Δv cost. Though that work solely addressed the ballistic case, other research by Anderson and Lo (2009, JGCD) in the Jupiter-Europa system demonstrated that invariant manifolds can play a role even in non-ballistic, low-thrust trajectory design. In that study, a Ganymede-to-Europa transfer optimized without any a priori knowledge of stable/unstable manifolds was shown to, in fact, “shadow” such manifolds quite closely.

While Anderson and Lo focused on the Jovian system, the most significant arena beyond GEO for future low-thrust missions is cislunar space. To better understand the role of invariant manifolds and MMRs in this regime, ESA’s SMART-1 mission provides an illustrative historical dataset. As the agency’s first lunar mission, SMART-1 utilized solar electric propulsion to reach the Moon. Launched in September 2003 and operating until November 2006, its trajectory followed a spiral out from GTO, eventually made a number of approaches at apogee relatively closer to the Moon, then passed through the lunar L1 libration point region in November 2004, and finally arrived in its final lunar orbit in February 2005. Visually, segments of the SMART-1 trajectory strongly resemble some of the ballistic resonant “highways” of Kumar et al. (2025), despite the presence of thrust. This resemblance is also reflected in archived mission data: during the final phase of its 331-orbit “Earth-escape” sequence, the perigees of orbits 330 and 331 had osculating semimajor axis values of 239,975.74 km and 241,560.12 km, respectively—values very close to the nominal 2:1 MMR semimajor axis value of 242,156.2 km.

To more rigorously determine the dynamical origin of these observations and provide insight for future low-thrust lunar trajectory design, this paper analyzes the final geocentric phase of SMART-1’s trajectory through the lens of lunar MMRs and stable/unstable manifolds. Because low-thrust propulsion renders the system non-conservative, the fixed-energy analysis used in standard planar circular restricted 3-body problem (PCR3BP) studies is insufficient. Instead, this study considers the spacecraft’s instantaneous Jacobi constant at each epoch, projecting its state onto 2D Poincaré sections generated at the corresponding energy levels. By overlaying the flown trajectory against the stable and unstable manifolds of resonant and L1 Lyapunov periodic orbits, we identify the specific resonant families that guided the spacecraft as well as the specific manifold branches and subsets used by the flown trajectory. Special attention is given to the trajectory segment bridging the 4:1 and 3:1 MMRs. In the ballistic PCR3BP, natural connections between these resonances are topologically blocked by stable orbits. This analysis characterizes how SMART-1 utilized low-thrust maneuvers to bridge the energetic and topological gaps between these resonant families, while still leveraging natural manifold structures.

Through this analysis of the SMART-1 case, we highlight the role of underlying natural mechanisms in guiding low-thrust trajectories to the Moon. This knowledge has the potential to reduce the time required for design and optimization of such trajectories by providing a clear, systematically replicable portrait of the pathways followed by such missions—a portrait which should provide useful, geometrically motivated initial guesses for the optimization process.