Embedded Software Engineering

Description

Embedded software is found in cell-phones, cars, robots, the electrical power grid, and anything else that isn't what we traditionally think of as a “computer” but needs software smarts built in to it. This course takes you beyond simply knowing how to program an embedded system, and looks at applying the principles of software engineering to embedded systems. Software engineering in general is about the application of a systematic, disciplined, quantifiable approach to the development, operation, and maintenance of software. The essential principles of software engineering apply to the development of embedded software, but the application of these principles sometimes differs due to the unique nature of embedded systems.

Developing software for an embedded system differs from developing desktop or server software because embedded systems must run on non-standard processors and custom PCBs, often must work without fail, must interact with the physical world through application-specific sensors and actuators, and are typically resource-constrained. Making the design of our software reliable and flexible enough to accommodate these needs involves careful use of abstraction and interfaces to break the design into understandable, testable, and adaptable modules. During the course we'll look at various principles and patterns for developing well-structured designs, as well as techniques for implementing those designs in code. To help understand how our modules should behave, and how the software will interact with the physical world, we'll look at the various models of computation typically found in embedded systems and where they are best applied. Knowing the primary models of computation in given design can guide our choice of execution model (cyclic, event-driven, time-driven, or some mix) as well as facilitating analysis and testing. Many embedded systems are faced with the problem of carrying out multiple tasks concurrently, so we'll also look at the various ways we can write code to do this, the pros and cons of different approaches, and design techniques for achieving robust concurrency.

The course includes two practical projects in which the ideas taught during lectures are applied to actual software design and implementation.