Abstract: Brain machine interfaces or neural prosthetics have the potential to restore movement to people with paralysis by bridging gaps in the nervous system with an artificial device. Cortical implants can record from hundreds of individual neurons in motor cortex. Machine learning techniques can be used to generate useful control signals from this neural activity. In recent years, there has been substantial progress improving these systems in several areas. Performance can now surpass typically used EMG control signals for artificial limbs, and animals can control computer cursors with brain activity 80-105% as well as they can with their native hand. There have also been substantial improvements in electrode longevity, due to signal processing techniques, which can enable devices to function across multiple years. Also, several integrated circuits have been developed to process many channels of neural activity with a very small device. However, significant challenges still remain to package these systems such that the active electronics can remain in the brain for many years, or to develop fully implantable neural interfaces that require no external wearable components. Finally, there is a large portion of the paralyzed community whose primary need is for finger and grasp control, which has not previously been attempted in a real time experiment. Currently, two human safety trials are being completed using multi-channel neural implants in motor cortex. If these challenges can be met, it is possible for this brain machine interfaces to become widely available for the treatment of paralysis. Bio: Cynthia Chestek is an Assistant Professor, Biomedical Engineering, at University of Michigan. She completed the B.S. and M.S. in Electrical Engineering at Case Western Reserve University, and the PhD at Stanford, also in Electrical Engineering. She completed a postdoc working on the Braingate2 human clinical trial for brain machine interfaces.