Introduction

Abstract

Embedded systems contain electronic computing elements (such as microcontrollers) with embedded computing program (such as firmware) and able to interact at least one element of surrounding. This chapter introduces the concept, provides basic block diagrams of such systems, and outlines history of these systems and future trends. Some key design metrics and challenges are also listed.

The scientists of today think deeply instead of clearly. One must be sane to think clearly, but one can think deeply and be quite insane.—Nikola Tesla, one of the greatest minds.

Exercise

Problem 1.1: Using a space heater as an example of an embedded system, draw a hardware block diagram of this system. Other than essential components, the system contains a manual keypad entry, a forced air heater, option for oscillating 120° range, and a display to show the set temperature and the sensed ambient temperature.

Problem 1.2: For an automated cashier machine at a store, draw a hardware block diagram. Other than essential parts, this system has a barcode scanner, a weighing machine, a card reader, a cash collector, a cash change dispenser, a cash reloading mechanism, a green/red light signaling mechanism to indicate status, a secure back-end connectivity with a financial transaction machine via Internet, a camera, a diagnostic port, and a touch sensitive display.

Problem 1.3: For an embedded system product with 3 years life cycle, a delayed product was launched after 6 months compared to on-time products. Compute the percentage potential revenue loos for the delayed product using a triangular revenue model approximation.

Problem 1.4: A company is planning to launch a new embedded system product. They estimate the product will have 4 years of life cycle. To avoid more than a quarter potential revenue loss, what is the maximum delay they can afford? Use a triangular revenue model approximation.

Problem 1.5: For an embedded system product with a market life of 60 months, compare the revenues for an on-time product and a delayed by 10 months product. Assume the revenue peak is $100 K. Use the triangular approximation model for the market window. Also, determine the percentage loss of revenue for the delayed product.

Problem 1.6: For an embedded system product, the NRE cost and unit cost are the following for the four technologies:

Technology	NRE expense	Unit cost
Semi-custom VLSI	$200,000	$5
ASIC	$50,000	$10
Programmable FPGA	$15,000	$20
Microcontroller	$10,000	$15

• (a) Calculate total per-unit cost for production volumes of 100, 1 k, 10 k, and 100 k units.

• (b) Plot these data from (a) in a single graph with log scale for per-product cost and draw piecewise linear lines for each technology. Then, determine the best choice of technologies for these production volumes (100, 1 k, 10 k, and 100 k units) to achieve the lowest per-product cost. Also plot total cost for these product volumes in a separate log-log graph.

• (c) From the per-product cost plot in the (b), estimate the range of production volumes for which each of these technologies is financially optimal.

• (d) List 3 other considerations in addition to per-product cost that might affect the choice of technology.

Problem 1.7: If a product unit cost is $40, and NRE cost is $10,000, determine the minimum quantity of units to be produced to keep per-unit cost below $50.