Significance: Robotic radiotherapy using extensively non-coplanar beams has been shown effective to significantly improve radiation therapy dosimetry, leading to improved treatment outcomes. However, the current implementation of this technique by CyberKnife is inefficient and not dosimetrically optimal. This has severely limited both the number of patients eligible for robotic radiotherapy and the achievable clinical outcome for those who have been treated. In order to overcome these limitations, a novel robotic radiotherapy system will be developed that can efficiently utilize the full potential of the non-coplanar delivery space to treat the majority of radiotherapy patients. Innovation: The proposed system is highly innovative in the following aspect: 1) Integrated beam orientation and fluence optimization. 2) Significantly more compact linac to allow posterior beams. 3) Flexible field sizes and MLC resolution to efficiently treat most target sizes. 4) Volumetric imaging and real-time IGRT will be implemented. This project is proposed to design the hardware and software platforms materializing such a robotic radiotherapy system. In order to reduce the gantry size, both the linac length and the distance between the source and the MLC need to be significantly reduced. A new 3 MV source has been designed to reduce linac length and provide the required dose rate for treatment. The physical MLC leaf thickness cannot be substantially thinner than 1 mm. To achieve a high MLC resolution at the treatment distance, a spacer is used in CyberKnife between the primary collimator and the MLC, increasing the gantry dimension. This proposed system will eliminate the spacer but vary the focus-to-tumor distances (FTD) to achieve desired field size and MLC resolution. This requires optimization in an enormous solution space, a capacity uniquely demonstrated by the 4? algorithm. Aims: 1a: Build the 3MV linac that can produce 800 cGy/min at 100 cm. 1b. Mount the linac on an industrial robot and test its mechanical robustness. 1c. Integrate a micro multi-leaf collimator (MLC). 2a. Develop a global optimal direct aperture solution for the static intensity modulated radiotherapy (IMRT). 2b. Develop a global volumetric modulated arc therapy (gVMAT) solution. 2c. Develop a navigation algorithm for the robot to travel and deliver the radiation efficiently. 3a. Perform safety and collision model test. 3b. Dosimetry end-to-end testing. 3c. QA test. Impact: Successfully achieving these three aims will provide a prototype to prove the feasibility of the versatile robotic system for radiotherapy. It will be scientifically and clinically significant, positioning the system well for further commercial development.