Design Metrics of Embedded Systems

1. Processing Power

A processor is chosen based on the necessary register width and the amount of processing power to complete the task.

2. Throughput or Performance

The execution time or throughput of the system is important design metric for an Embedded System design. Performance is measured by the system’s instruction execution time. Higher performance equates to shorter execution times.

For example: The processing of voice signals between the antenna and speaker in a mobile phone in 0.1s demonstrates phone performance. The system may need to process a large amount of data quickly.

3. Response

The system must respond immediately to the changing occurrences is an essential design metric for an Embedded Systems.

4. Memory

The best estimation of the amount of memory needed must be made during hardware design, along with space for expansion.

5. Power consumption

Systems typically operate on batteries, therefore power-saving strategies must be considered throughout software and hardware design.

6. Number of units

 The trade-off between manufacturing cost and development cost will be determined by the anticipated volume of units produced and sold.

7. Expected life-time

The length of the system’s estimated lifespan will affect design choices including component selection and system development costs.

8. Program Installation

Software installation on the embedded device requires specialized development tools.

9. Testability and Debug ability

Without a keyboard and standard display, it will be challenging to set up test circumstances and equipment and difficult to figure out what is wrong with the software.

10. Reliability

The developed system must always produce the output for which it was intended.

11. Power Dissipation

 This is a crucial feature for battery-powered systems. Examples include a mobile phone or a digital camera, where if power dissipation is low, the battery needs to be recharged less frequently.

12. Unit cost

The price of producing each unit of the system, excluding NRE costs.

13. NRE cost (Non-Recurring Engineering cost)

Expense incurred in creating the system. The word “nonrecurring” refers to the fact that, once the system is designed, any number of units can be produced without incurring further design costs.

14. Size

The amount of physical space needed by the system, frequently expressed in bytes for software and transistors or gates for hardware.

15. Flexibility

The ability to modify a system’s functionality without paying a high NRE cost. Software is frequently thought of as being incredibly flexible. Flexibility in design makes it possible to eventually produce more advanced versions of the product without incurring major technical costs. Adding more features to software is one form of software improvement.

16. Maintainability

Focuses on providing end-user or client assistance and maintenance in the event of technical problems and product failure. Less maintenance is associated with a system that is more reliable. Probabilities of failure and malfunction decrease as system reliability rises.

17. Time-to-market

The length of time needed to design and construct the system before it could be marketed to customers. Design time, manufacturing time, and testing time are the key factors. There could be several developers of items in the same category in the embedded market (like mobile phones, portable media players etc.). If you release a new product with a lengthy time to market, a competitor may exploit it with their own offering.

18. Time-to-prototype

The time required to create a functioning system, which could be bigger or cost more than the final system implementation but can be used to test the system’s accuracy and usefulness and to improve its operation. The actual anticipated development time can be reduced by accelerating the prototype’s development.

19. Correctness

Designers need to be sure they’ve implemented the system’s functionality correctly. Throughout the system design phase, designer may test the functionality, and can add test circuitry to make sure production went smoothly.

20. Safety

A measure of the likelihood that a system won’t be harmful. It addresses potential harms to users, the general public, and the environment that may result from embedded system failure or from the release of radioactive or hazardous materials from embedded devices. Product engineering requires safety analysis to assess the potential harm and choose the best course of action.