Following the invention of the atomic force microscopy, AFM has revolutionised not only surface sciences but also biological sciences more recently. From being used as a microscope with atomic resolution generating 3D topography of surfaces to being used as a micro and nanomanipulation tool to manipulate single atoms; from studying the dynamics and mechanical properties of different cell types and molecules to being used in force spectroscopy and studying cell stiffness to detect cancer. Nonetheless, a majority of such work has occurred for single-scale manipulation with one probe. There have been some work in the past two decades to achieve parallel manipulation but none of them have been able to achieve independent 3D actuation of the probes critical for targeted biomanipulation. We propose a new type of parallel architecture based microrobotic actuator that integrates arrays of microneedles with independent 3D mobility. We focus on the design of the parallel architecture and more specifically on a single-unit actuator (4SA and 3SA microrobot). The parallel architecture is designed as a hexagonal shaped structure to enable accommodation of multiple 3SA microrobots in a small space with efficiently laid electrical interconnects. It is found that the 3SA microrobot performs better in terms of motion performance and integration in the parallel architecture. The range of design dimensions are conceptualised and the design is analysed using extensive analytical and finite element analyses models. The 3SA microrobot can achieve a displacement of up to 72 μm in-plane at 160 V and 7 μm in out-of-plane at 35 V. We have also successfully demonstrated the vertical motion using a parallel-plate actuator arrangement of a long standing silicon tower underneath a microstage. Our first fabricated 4SA microrobot achieves an in-plane motion of up to 10 μm at 120 V and more than 0.5 μm at 600 V. We also propose two new fabrication process sequences designed to fabricate the microrobot and introduce a new blind feedback mechanism for achieving vertical biomanipulation by detecting the change in voltage-displacement plot as signature for manipulation. Our PID control mechanism utilises visual feedback for in-plane alignment and blind feedback for out-of-plane alignment of the microneedles. We have successfully demonstrated the feasibility of our blind feedback system by designing a macro-manipulation experiment.