We are entering a new era of synoptic surveys where alerts for young transients are increasing at a rate faster than they can be followed up. The bottle-neck is related to spectroscopic follow-up, with the problem getting worse as surveys get wider and deeper, moving towards the LSST era. To remedy this dearth of spectral follow-up we are building a fast-slewing, fully robotic 4m class telescope on La Palma, to be known as the New Robotic Telescope. The workhorse instrument is to be a medium resolution, high-throughput IFU spectrograph. Other first-light instruments are to be a fast readout camera, a wide-field camera and a polarimeter. The aim of being on target 30 seconds following an alert will allow us to explore a previously difficult to observe parameter space. Many areas of transient physics will be positively impacted by fast and frequent spectral follow-up of transients at early times. For example, the relative frequency of the ever-increasing zoo of exotic supernova subtypes can be addressed and their unusual environments probed. The new facility will work in parallel with the synoptic surveys by not only providing object classification, thanks to the broad spectral range to be covered, but will also better observe the high velocity features in early time SN Ia spectra. Early spectral observations will help to pin down contested issues like deflagration and detonation models; the single/double progenitor question and spectral evolution of rare supernova subclasses. Other key science drivers include gravitational wave counterpart follow-up and gamma-ray bursts. The advantages of having a fully robotic observatory is that both down-time and time-to-target are significantly reduced.