In this paper we report on our ongoing research on wearable active vision, where we have iteratively prototyped a Wearable Visual Robot a body mounted robot for which the main sensor is a camera. Two main areas have been studied: robot design and visual algorithms. In the design stage, we have analysed sensor placement through the computation of the field of view and body motion using a 3D model of the human form. A design methodology for the robot morphology was developed with the help of an optimisation algorithm based on the Pareto front. The wearability of the device has progressed over several iterations as have the sensor and control architectures. In terms of visual algorithms, we have studied methods of visual tracking fused with inertial sensors, real-time template tracking, human head pose recovery and more recently real-time simultaneous ego-localisation and autonomous 3D map building. Our main long-term application areas are enhanced remote collaboration and autonomous wearable assistants that use vision.