This work presents a robot prototype designed and built for a new aided fruit-harvesting strategy in highly unstructured environments, involving human-machine task distribution. The operator drives the robotic harvester and performs the detection of fruits by means of a laser rangefinder, the computer performs the precise location of the fruits, computes adequate picking sequences and controls the motion of all the mechanical components (picking arm and gripper-curler). Throughout this work, the specific design of every module of the robotized fruit harvester is presented. The harvester has been built and laboratory tests with artificial trees were conducted to check rangefinder's localization accuracy and dependence on external conditions, harvesting arm's velocity, positioning accuracy and repeatibility; and gripper-cutter performance. Results show excellent range-finder and harvesting arm operation while a bottleneck is detected in gripper-cutter performance. Some figures showing overall performance are given.