Abstract
Background/Objectives: To determine the optimal site preparation and placement protocols for immediate implant positioning in robot-assisted surgeries. Methods: In vitro models of immediate and healed extraction sockets were created using 3D printing. A robotic system was used for implant site preparation and implant placement. The implant surgeries were allocated into eight experimental groups using 12 printed models in total. Each model incorporated two implant sites, an immediate site (tooth 21) and a healed site (tooth 26), resulting in 24 implants overall. With 3 implants assigned to each group, the 24 implant placements were evenly distributed across the 8 groups. For each group, the lateral force experienced during surgery was recorded by the haptic sensor on the robotic arm, and implant positional deviations were assessed by superimposing post-surgical CBCT images with the virtual implant planning. Results: Healed sites showed significantly higher accuracy than immediate sites, with reduced platform and apical deviations (p < 0.001) and markedly lower lateral force experienced by drills. In fully guided procedures, thread tapping greatly improved accuracy in immediate sites but had limited benefit in healed sites. Compared with partially guided workflows, fully guided rCAIS markedly enhanced accuracy in immediate sites (≈0.8 mm reduction in platform/apical deviation, p < 0.001), while no meaningful differences were observed in healed sites. Fully guided protocols also reduced insertion force in healed sites. Conclusions: Immediate sites showed lower implant positional accuracy and experienced higher lateral forces during surgery than healed sites. In immediate sites, thread tapping and fully guided rCAIS improved placement accuracy.