Differences between Application and Embedded Testing

Differences between Application and Embedded Testing

Embeddedness:

Most embedded software is, well embedded. So they do not generally have a key board, hard disk and monitor attached to them. In that case, special mechanism needs to be created to validate the software and the complete system before actual usage.

Real-time behaviour:

Most of the embedded systems are real-time. It is important that output received from the system needs to be correct and also that it arrives within a specific time. Later response may be is correct.

Difficult to simulate

Due to the conditions under which embedded systems are used, the simulation of its actual environment may be expensive, difficult, or dangerous. A simple embedded device such as an electric or smoke detector creates problems since the validation system needs to circumvent the actual sensor and send a signal directly to the software executed because of the signal.

Difficulty in seeing the output

Since embedded systems are usually connected to devices, their generated outputs are not in the form of a message on the screen, but they may be a command to handle the device or write something in memory. So the test suite needs to explore the internal parts of the embedded system in order to verify if the desired operation was performed.

No down time

Unlike application software, many embedded systems are expected to run continuously. This poses its own problems, and they are very difficult to detect during validation. Problems such as memory leaks, recovery from incorrect states, hardware malfunction, etc. usually are difficult to simulate.