I decided that the targeting/sensors discussion in the CU proposal thread was starting to take it over a bit, and I ended up with enough ideas on the subject to put together something approaching a coherent system. So, here it is (TL;DR first): EDIT: Here's a high-level system overview, inspired by @Hicks42 : The sensor system has a sensitivity threshold below which it doesn't see a signal (which also determines FoW reveal). Processing a raw signal can increase its level enough to be considered detected, but takes computational effort (CU hours) An object's signal level increases with its effective area and the detector's power output, and decreases (quickly) with its distance from the detector Ambient noise levels and jamming equipment constitute a background noise level that can mask weak signals even with post-processing contributes to a false positive/negative rate Active IFF transponders make ships easier to detect since the IFF signal only has to travel one way and thus its level doesn't fall as rapidly with distance Active stealth mode works to decrease a ship's effective area, and consumes power and CUs to do so. The degree to which a signal is above the sensor threshold modifies the accuracy for AI/NPC-aimed weapons. Next, a little bit of background; there are 5 main aspects to detection/targeting: Size of target vs distance to target: The closer you are to a target and the larger it is, the better you can see it, all things being equal (reflected power from a target goes as A/R^4, where A is the area of the target, and R is the distance to the target). Emitted wavelength: The shorter the wavelength used, the longer your diffraction-limited range. Emitted Power: The more power you can put into your active scanners, the better you can see your target (more reflected power). Detector gain: The higher your gain, the smaller the signal you can see. Signal processing ability: The more compute power you can direct to processing your sensor data, the better you can resolve your target from the surrounding noise. A substantial improvement in any one of these areas can, to some extent, compensate for deficiencies in the other areas. The overall goal here is a system that can provide an accuracy and/or damage multiplier for combat (including the effects of stealth), as well as provide a way to gather data from enemy ships and the environment for research purposes. In order to simplify things a bit, I will assume that wavelength and gain are subsumed into quantity that is related to the quality or tier of the sensor system and which gives the minimum signal threshold; let's call it T for now. Let's call the background noise level N, and the power level of enemy jammers be J. Finally, clever signal processing algorithms (S) can help raise a signal above the minimum signal threshold. Now, we know that the reflected power should be proportional to A/R^4, so we have Pr = Pt*A/R^4, where Pr is the reflected power, Pt is the transmitted power, and A and R are as above. Putting this all together, the signal (D) detected by the sensor system for a certain object at a certain distance in an environment with a particular noise/jamming level is the following: D = Pt*A/R^4*S - (N+J) Then the condition for detection is given by: D > T Next, we need to figure out the devices in this system. I propose 2 types: Emitter/detector devices, which could be the existing radar and antenna deco devices; would consume large amounts of power Sensor system devices, which process the signals received by sensor system; would consume large amounts of CUs The core has a compact sensor system built into it, so that a structure always has some sensing capabilities; adding one or more of each of these device types would expand its capabilities. Passive stealth would be accomplished by flying far from enemy structures in noisy environments (like a nebula); active stealth would be using the sensor system to reduce the effective area of the ship (which would consume considerable power and CUs). As far as combat is concerned, if a target can be detected, one straightforward option would be to take the quantity log(D/T) * 0.5, capped at 2, and multiply that to accuracy calculations for turrets and homing weapons. At the extreme edge of sensor range, there would be an accuracy penalty which would go away once the target was at about 60% of the max sensor range, and the maximum improvement would occur at about 35% of the max sensor range. I haven't fully worked out how data collection ought to work, so I'll leave that for the next post.