Due to the shortage of medical personnel, no ultrasound screening campaigns on the population are available for identifying dangerous and asymptomatic vascular diseases. Therefore, we propose the development of an autonomous robotic system that could enable the execution of standardized ultrasound procedures for extracting relevant vascular indicators. The robot moves the ultrasound probe on the patient tissues applying the proper forces for generating a good ultrasound view. The image is then processed by a deep neural network that segments the vascular structures and provides the final indicators. The system implements three different use cases: the detection of abdominal aortic aneurysm, the estimation of central venous pressure from jugular ultrasound and the execution of the compression ultrasound to identify deep vein thrombosis.