Mobile robot controller

1
Mobile robot controller

• Group 2

2
Group 2

• Leif Hamang Bru
• Lisa W. Eriksen
• Kay Are Ulvestad
• Gunnar René Øie

3
Introduction

• Project problem
• What we should do
• The quality attribute in focus was availability

4
Blackboard architectural pattern
5
Attributes realised in architecture

• Availability
• Redundant blackboards
• Voting
• Possibility for redundant crewmembers

6
We have learned

• We should have learned more about the robot
  before setting our goals, and taken a simpler
  approach.
• Got some more documentation practice.
• The ATAM process should be used with larger
  teams.
• Using software architecture may be hard if the
  system is small and highly interconnected.

7
Findings/achievements

• ATAM works as a structured way to give feedback.
  Feedback was valuable, to improve explanation of
  architectural choices.
• Quality requirements are sacrificed to achieve
  functionality
• We were able to make all the modules needed to
  solve the problem, but didnt finish putting them
  together. Why Discrete movement problem.

8
Demonstration

• Strategy
• Build internal map
• Map-matching
• Cover map
• Move all balls