Flexible Delta robots show significant promise for use in a wide array of manipulation tasks. They are simple to design and manufacture, and they maintain a high level of repeatability and precision in open loop control. This thesis analyzes the kinematic properties of flexible Delta robots and addresses questions about the control and design of Delta-based manipulators. We develop an analytical model for the kinematics of a rigid Delta robot with approximated universal joints, which is then used to compare the kinematics of flexible and rigid Delta robots. We develop an algorithm to detect and prevent collisions between adjacent Delta robots. We model the flexible Delta kinematics using neural networks, obtaining ground truth from computer vision, and explore the generalizability and repeatability of the learned kinematics. Finally, we obtain initial force profile results, and identify questions to be addressed in future work.