Design of Embedded and Intelligent Systems, 15 credits

Introduction

The main part of this course is a organized in the form of a system development project. This project is to be performed in groups of 4 students. To support the project a number of seminars, lectures and labs that provide knowledge and skills are given and compulsory. Each group of students must include members with skills from the three possible specializations (embedded, intelligent and communication) since they are all necessary to carry out the project task. The goal of the project is is to design and develop a system solution to a problem defined in a project specification that all groups have to follow. Each group creates their own solution to the problem and the different solutions are then compared, discussed and examined as a final part of the course.

Course Organization

The project is described in a Project Specification. (pdf, 52.3 kB) There is also other information and material for the project, that will be made available during the project.

Seminars/lecture areas:

1. Activities in early phases of a design project (Tony L 6 hours)
Project task definition and project group formation. The problem understanding, design space limitations, and possible solution approaches. Capturing, analyzes and understanding of market, needs and requirements of a system and its functions and characteristics. How to work as a group in a design and development project. How describe and model the systems architecture in terms of physical and functional elements. Use of UML and similar notations to model a system and its parts, functions and behaviors. Design space exploration and trade-off decisions. Presentation/reading material 1. (pdf, 606 kB)

2.  Programming of Embedded Systems (Nicholas N 6 hours)
Software development environment, tools and methods for programming, compilation and debugging of embedded software. Use of support from real-time operating system. Praxis regarding software modularization and integration.
Presentation/reading material 2.

3.  Navigation, Trajectory Planning and Control of Vehicles (Ulf H 4 hours)
Planning of trajectories, directions, speed and dynamic changes. Correction of trajectories, speed, direction and position feedback using camera measurements.
Effects of delay in the control loop.  Presentation/reading material 3 (pdf, 572.6 kB).

4. Combination of Sensor Information (Antanas V 2 hours)
About  different types of sensors and how they can be combined to gain accuracy by fusion of information from several sources with different quality. Presentation/reading material 4 (pdf, 541.4 kB).

5. Wireless Communication (Urban B 2 hours)
Sharing of radio channel capacity (frequencies, time slots etc.). Coordination of multiple access attempts. Frequency hopping, acknowledge and resending to improve robustness. Presentation/reading material 5 (pdf, 181.8 kB).

6. Image Processing for Surveillance and Control (Josef B 4 hours)
Identification of color and pattern coded objects. Detection of image objects and their direction and position vectors. Presentation/reading material 6.

Lab exercises (mini-projects):

L1 Programming of drivers and support functions related to the sensors, actuators and communication devices to be used in the project (Nicholas 4 hours). Lab material L1.

L2 Use of timers and a simple real-time operating system for task scheduling and handling of deadlines , periods and shared resources (Nicholas 4 hours). Lab material L2.

L3 Drive of vehicle by help of program control from one position to another with feedforward using a kinematic robot model and with feedback from camera measurements.
The software Sysquake can be used for studying spline tracking.
Sysquake file for robot control using splines: robot.sq.
Sysquake file for defining slines: spline.sq.
(Björn Å/Ulf H, 4 hours).

L4 Fusion of signal data from simple sensors on feature level and decision level. (Jens L and Anita S, 4 hours). Lab material L4. (zip, 254.5 kB)

L5 Use of the 2.4 Ghz radio module for wireless communication based on coordinated sharing of the radio spectrum and use of a simple Master-Slave protocol (Tommy S/Björn N, 4 hours). Lab material L5.

L6 Acquisition of image from camera via MATLAB and HTTP command. Calibration and correction of the image. Classification of objects in image. Estimation of the objects position and direction. Handling of color and luminance information. Use of image sequences to estimate speed and direction etc. Subtraction and addition of image information, up/down sampling, filtering, pattern recognition, thresholds, smallest distance, etc. (Josef B, 4 hours). Lab material L6.

Examination Criteria

The project is performed following a project control process dividing the project into a set of tollgates. At these tollgates each group has to report about their achieved results both orally and by written documentation. In latter phases the reporting also will require practical running demonstration of the system solutions.

Project Tollgates

Tollgate deliveries such as oral presentations, written documentation and live demonstrations are all part of the grade given.
 
EX-1 (w38): Refined definition of the problem and important limitations documented in a system requirement specification defining the systems functional and characteristics requirements.
 
EX-2 (w42): System architecture description including component and task partitioning with requirements on each component/task. Common precedence and collision avoidance rules defined and agreed upon by all groups and documented in the description.
 
EX-3 (w46): Proposed, implemented and tested component and task level solutions as well as arguments for different design trade-offs documented in written report.
 
EX-4 (w50): System integration, system test report and preliminary demonstration of the system.
        a) A single vehicle can be driven between several positions.
        b) The vehicle can avoid static hinders placed on the play ground by the examiner.

EX-5 (w02): Final test demonstration and examination including all groups and their vehicles, followed up by discussions and reflections about achieved experiences and results. The vehicle can be driven between several positions on the play ground together with the other vehicles driving simultaneously and avoid collisions with these in an intelligent way.

Grade 3 The group has delivered the (reasonably commented) implemented-software including training data if any, the final report, and the final presentation slides on a CD and the individual can explain and also demonstrate how the system works in an acceptable way.

Grade 4 In addition the project task is made in a nice way and the group/individual has been able to perform the project task in a disciplined, self managed way without too much external help.

Grade 5 In addition the group/individual can motivate selected solutions and argue for pros and cons compared with other possible solutions.

The grade given is weighted in relation to achieved demo results in terms of speed, precision, safety and dependability, not only for the own vehicle but more important - as a contributor to the whole system.

Links

Tony's homepage