Through my whole career, the quest has been to build better and better instruments to collect more photons. For example, the Cosmic Origins Spectrograph (COS) on Hubble, is able to collect information more than 20 times more efficiently than the older high-resolution spectrograph, Space Telescope Imaging Spectrograph (STIS). This is because all design trades were made in favor of maximum sensitivity. As a result, in other aspects, such as imaging capability and spectral resolution, it is less powerful than STIS.
Astronomers always want to look at fainter objects, more distant objects, and they want to get more precise detail about that object or that gas they’re trying to measure. In spectroscopic terms, that usually means obtaining better spectral resolution, determining exactly how hot the gas is, learning how fast it is moving, finding the exact ratio of carbon to hydrogen in this gas, and so forth. Those are all things you can get from ultraviolet spectroscopy, and the better your instrument, the better you can measure such things. But, you also need enough signal from the object to measure those items. If you have a perfect instrument but only get one photon, you don't really get any information. So, you have to build something that will allow you to collect a lot of photons, because otherwise that will limit the amount of information you can extract.
Astronomers used to ask the Hubble scientists for three days of observation with Hubble to study a distant galaxy. The standard answer was that three days was too much time: it’s 1% of a whole year’s observing time. Now, that same observation might take three hours, or even one hour, and then the project can be approved. What everyone has been waiting for over a span of 20 years has finally materialized, and that didn’t happen because we put up a bigger telescope. It’s the exact same Hubble mirror collecting the light. It happened because we put forth a design that optimized sensitivity over all other considerations.
For the future, considering that new instruments will be possible, we would like to improve the detectors that collect the light by improving their efficiency and their ability to precisely locate each photon in an object's spectrum as it arrives (spectral resolution). We also need to improve the efficiency of mirrors so they are more readily usable in the design of ultraviolet spectrographs. These are the types of things that I’ve been working on beyond working on current spectrographs.