If you see a human body, then one of the most distinctive parts that makes us able to be different than most other species is the arm. It is the part of the body that we use to do most of the work.
The human arm is a very complex mechanism of joints and muscles that work in tandem to give it the dexterity that we know it for. Take an example of our shoulder joint. If you pay attention, then you will notice that it has the ability to move up and down, right and left, and even rotate on its own axis, and all this while it just has one single joint, which we know as a ball joint.
When we talk about a robotic arm on a robot, we are undoubtedly talking about a complex arrangement of actuators with the body, otherwise known as a chassis, to get the desired motion in a three-dimensional space.
Now, let's understand some of the basic parts of any robotic arm. The first parts are the actuators. We can use motors to control the robotic arm; however, as we have studied earlier, using the motors we have used before will not be the ideal solution for it as it cannot hold its position neither does it have a feedback mechanism. So we are left with only one option, that is, to use servo motors. As we know, they have a handful of torque and have the ability to know where it is and to hold its position for as long as we want.
The second part of the robot is the chassis, that is, the part that holds all the motors together and provides structural support to the robot. This has to be made in such a way that it provides motion in all the desirable axis to any given joint. This is important as a single servo can only provide motion in one single axis. However, there are multiple places in which complex arrangement can be used to make the robot traverse in multiple axes. Also, the chassis should be rigid, which is extremely important. As we all know, all the material on this planet have certain level of flexibility. Also, the construction of the material depends on how noncompliant the material would be. This serves a very important purpose of repeatability.
Now, what is repeatability? As you might have seen in industries or any manufacturing units, the robots are installed and they do the same task over and over again. This is possible as the robots are programmed to perform a specific set of functions under specific circumstances. Now, let's say that the chassis of the robot is not rigid. In such a case, even if the servos are 100% precise and get to the exact same position over and over again, still the robot may actually differ from its actual goal position. This happens as there may be some flexibility in the chassis, which is why the final position may differ. Hence, a right chassis is a must. It becomes even more important when we are talking about large robots, as even the slightest of deformation can lead to a very large change in the final position of the arm.
One very common terminology which we use while talking about the robot arm is the end effector. This is basically the end of the robot arm, which will be doing all the final work for us. End effector in the case of a real human arm can be considered the hand. This is at the top of the arm and all the movement of the arm is basically to articulate the position of the hand in a three-dimensional space. Also, it is the hand that picks up the objects or does the necessary physical action. Hence, the term end effector.
Now, as the robotic arm is moving in a three dimensional space, it becomes a real big problem to define the axis in which the motion is happening. Hence, instead of using the axis to define the motion, we generally use the type of motion being performed, which gives us a realistic idea of what the motion is and in which axis it may be on. To analyze the motion, we use the concept of Yaw Pitch and Roll (YPR).
The preceding diagram will clear most of the doubts about YPR. This concept is generally used in aircrafts; however, it is an essential part of a robotic hand as well. So, as you can see from the preceding figure, when the nose of the plane goes up or down, it will be considered as pitch movement. Similarly, if the aircraft changes the heading, then the Yaw can be considered to change accordingly—the Yaw is nothing but the movement of aircraft in the y axis. Finally, we have something called Roll. It is used to understand the angel of rotation. As you can see, all these three entities are independent to each other and chasing any of it will not have any effect on the other. This concept is also useful as no matter what the orientation of the aircraft is, the YPR would still be unchanged and very much understandable. Hence, we take this concept straight from the aircraft directly to our robots.
Finally, how can we forget about the processing unit? It is the unit that commands all the actuators and does the coordination and the decision making. This processing unit in our case is Raspberry Pi, which will command all the actuators. All of these preceding components make up a robotic arm.