\fExercise 2: Detection and Classication of Motion In this exercise, you will write a sketch that can distinguish between a circular motion and a reciprocating motion {i.e. back-and-forth linear motion), and possibly computej'detect other aspects of these motions [like direction, frequency, etc.], using an accelerometer. There are no additional experiments required beyond the checkoff. However, this part is largely an experimental design of a detection algorithm. You will be performing a lot of experiments, gathering and analyzing data, and possibly trying several different approaches. Document them while you're designing the algorithm [even if an approach didn't work} and present them in your report in detail. The movement of the accelerometer will only be limited to the horizontal plane. If your accelerometer is placed on the desk with Z-axis pointing upwards, then the X and Y-axes are the only directions of interest. For reciprocating motion, you will move your breadboard with the accelerometer attached to it back and forth approximately in a straight line. This line can be along X-axis, along Y-axis, or at any arbitrary angle. The motion can be slow, medium or fast. For circular movements, you will move the breadboard along a small circle (clockwise or counter-clockwise], preferably at a steady angular velocity. While you're moving the breadboard, the X and Y axes should always point to the same direction leg. east and north, respectively). For both these motions, since you will be moving the breadboard manually, the motion will not be perfectly circular or reciprocating {i.e. the signals will be noisy]. However, for a somewhat approximate linear or circular movement, it's possible to distinguish between them by processing the acceleration data properly. When there is no movement, your system should be able to detect that as well. However, when there is motion that are neither circular nor linear, or the direction andfor speed changes abruptly, it's possible that your system will give false decisions. You will develop your algorithm mainly by observing the trend in accelerations in X and Y directions for different motions. You can start by just manually observing the data in the serial plotter and trying to find patterns. You can also oompute certain metrics like average, variance, correlation, etc. and see how different motions, speeds, directions, etc. affect these metrics. Once you identify certain attributes of the acceleration signal for a specific type of motion, you have to translate that into an algorithm so that a sketch can make decisions. Whether and how you implement any sort of calibration at the beginning is up to you. Your sketch will periodically ask the user to move the breadboard, record and process the data, and display its decision. An example task flow can be as follows: 1) At startup, the sketch will perform calibration, if any, and may ask the user to do certain things. It will then repeat steps 2 to Sin a loop. 2) Ask the user to start movement within 5 seconds. During these 5 seconds, the sketch will do nothing; this is just to allow the user some time to initiate a motion. 3) Record andfor process the data for 10 seconds. During these 10 seconds the user will continue the movement steadily. 4) Ask the user to stop movement. 5) Process the data and run the detection algorithm, and display the results in the serial monitor. Checkoff 5.2 [25 aims]: Let Your TA/instructor know when you're ready to demonstrate. Here are the basic requirements of the sketch: If motion is circular, it should always display "Circular Motion". If motion is reciprocating, it should always display "Reciprocating Motion". If there is no motion, it should always display "No Movement". If there is an arbitrary motion that is neither circular nor reciprocating, it's okay to falsely output "Circular Motion\