Imagine you are feeling very hot. You are 'sensing' the room temperature. Your brain processes this sensation and directs you to switch on the AC. Your brain has,in a sense, designed an ‘algorithm’ for you. It came up with a solution for you. The AC, here, is the 'actuator'. It helps you get to the desired room temperature. Now, after sometime you realise that the room temperature has dropped really low. You ‘sense’ this again and as a result of this ‘feedback’ your brain asks you increase the temperature. Notice, this process has gone around in a closed loop. Your feedback of the output made you respond again and adjust the actuator. <br \>

What we did above is to ‘control’. However, a control system that is designed has machines do all this job. Broadly, control has three components- ‘sensors’ to assess the current state of the system, ‘processor’ to execute the control law, which is an algorithm based on the requirements (that’s what your brain did!), and ‘actuators’ as tools to carry out the controlling action. <br \>

Specifically for a satellite, once reference and body frames are defined, the next goal is to align these frames. This is what we try to 'control'. This reference, for instance, could be the orbit frame based on the satellite requirements. Remember, the orbit frame has to be defined at every point of orbit. In this case we would require sensors like the gps sensor to get the location and velocity of satellite and accordingly define the orbit frame. <br \>

A satellite ,usually, has other sensors too. These measure some physical quantities in the body frame. The same quantities are calculated for in the reference frame from some standard scientific models. We compare how these vectors are in the two frames to know how the body frame(satellite) is oriented with respect to the reference frame and hence deduce the attitude. For example, in Pratham we had sun sensors and magnetometers fixed to its body to measure the sun vector and magnetic field vectors respectively. <br \>