Traditional additive manufacturing is constrained by the workspace of the printer, i.e.
printers can only print objects within the printer's boundary. Mobile 3D printing is
developed here to fabricate large-scale objects that extend beyond a printer's workspace.
Mobile 3D printing uses a small-size robotic system to build large objects by connecting
multiple small segments. A possible example application for this is additive construction
on extraterrestrial surfaces, using locally sourced material, to minimize the overall need
for equipment and materials launched from Earth.
The system is equipped with both a laser total station (range and bearing sensor) and
3D scanner; measurements from these two sensors are fused to overcome the de�ciency
of each individual sensor. An Extended Kalman Filter (EKF) based Simultaneous Localization
And Mapping (SLAM) algorithm is implemented in order to align neighboring
segments. A representation for planar patches of the model being printed, with each
patch represented by 2 angles for the normal vector plus a 3D point on the patch, is
proposed and shown to be particularly suited for this type of task.
The system achieves sub-millimeter geometric accuracy and avoids the SLAM inconsistency
problem for well beyond the bounds of odometry error that could be expected to
be encountered in practice.