How to build better remote sensors for better science?

Today, NASA published a news about it's Innovative Research Grant activity that conducts engineering analysis to demonstrate the feasibility and advantages of applying a breakthrough remote sensor calibration concept to a wide range of future NASA remote sensor science missions.

The proposed concept enables future scientific and climate-class remote sensors to achieve the very low measurement uncertainty needed to accurately trend the earth radiation balance and to isolate cause/effect through highly accurate, spectrally resolved measurements in the UV/VIS/NIR from space. Such a breakthrough in remote sensor measurement accuracy is needed to obtain reliable measurements from space capable of trending decade long changes in earth climate and then associating those observed changes with root cause.

I was wondering if CEans here can discuss how they can better the existing remote sensors that engineering students use in projects. What do you guys think about this?