Emerging machine-learning tools have the potential to replace the limited capacity of human actors to monitor security video. New surveillance tools known as video analytics (VA) have the ability to perform the comprehensive monitoring of what is being recorded by public surveillance cameras and immediately alert personnel when various hazardous objects (for example, a raised gun) or movements (for example, a fall from a bridge) are viewed by a camera. The decision on whether and how to respond to VA alerts remains with human actors. On a parallel track with VA, another technology is refining ways to merge multiple data streams intelligently. This technology could automatically aim cameras at an officer when biometric sensors signal that the officer is under stress or assault, or this technology could trigger a video zoom on a license plate when a nearby camera detects activity consistent with a veering or suddenly speeding vehicle. This technology is called sensor fusion (SF). In helping to develop a responsible path forward for VA/SF research, a 15-member panel of policing and security experts convened by NIJ has created a VA/SF “investment roadmap” for technology development that identifies needs for innovation. The panel has concluded that VA and SF are “extremely promising technologies for improving public safety.“ Potential uses are noted for real-time monitoring, forensics, auto-reporting, performance monitoring, and officer safety and health. This report lists crimes in progress noted by the panel that should be recognizable through applied VA/SF. Crime precursors that should be noted by VA/SF are also listed, along with restrictions on the use of VA/SF.