9.4.1 Unit Testing, Integration Testing, System Testing, Acceptance Testing

Software testing happens at different levels, like checking different parts of a robot:

Unit Testing: This is like checking if a single motor or a single sensor on your robot works perfectly by itself. In software, unit testing means testing the smallest individual parts of the code (like a single function or a tiny piece of a program) to make sure each part works as expected. It helps catch small mistakes early.
Integration Testing: After checking individual parts, you need to see if they work together. This is like checking if the robot's arm can connect to its body and move smoothly when both parts are working together. Integration testing checks if different parts or 'modules' of the software work correctly when they are combined.
System Testing: Now, you check the whole robot as a complete system. Can it move around the room, pick up an object, and avoid obstacles? System testing checks the entire software system to make sure all its parts work together as a whole, and that it meets all the overall requirements. It's testing the complete product.
Acceptance Testing: Finally, you ask the person who ordered the robot to try it out. Do they 'accept' it? Can it deliver pizzas the way they wanted? Acceptance testing is done by the actual users or customers to make sure the software meets their needs and is ready for real-world use. If they are happy, the software is 'accepted'.

Bibliography:

Myers, Glenford J., Sandler, Tom, and Badgett, Corey. The Art of Software Testing. Wiley, 2011.