IBM Service Oriented Modeling and Architecture TM (SOMA) Licensed Materials - Property of IBM © Copyright IBM Corporation 2009. All rights reserved. Introduction.

IBM Service Oriented Modeling and Architecture TM (SOMA)

Licensed Materials - Property of IBM © Copyright IBM Corporation 2009. All rights reserved.

Introduction to Robotics

Dr. Ali Arsanjani

2

Robotics Engineering Academy

© Robotics Engineering Academy, 2009. All rights reserved.

PART 1: THE ENGINEERING PROCESS

3



The Engineering Process: Part 1 : the Cycle of Progress

Transcend

Plan and Design

Prototype &

Test

Integrate

& Test

Record & Reflect

hardware

software

GOAL

Capabilities & Mechanisms;Strategies for using the above

Add the sub-system you tested to the larger structure

Record the results and reflect on what adjustments you need to make

Decide if you have reached your goal, if so, choose another goal and its capabilities

4



Plan and Design

Scope the effort; – how much will you undertake in this iteration?

Plan : decide who will do what in what amount of time!

– Roles, tasks, durations, finances

Design the capabilities and the system or sub-system that those capabilities form

5



The capabilities can be grouped into “systems”

Collect balls

Deploy balls

Trigger the Ball Chute

Start Bush mechanism

Reverse Bush mechanism

Lower ramp

Start motors

Raise Loader

Deployment systemCollection system

How do they interface? (connect)

A system: a cohesive group of capabilities working together towards a goal.

6



Prototype and test

Build a prototype for the set of capabilities (sub-system) you have chosen to focus on

Test the prototype

Change it until you are meeting your goals and objectives

7



Integrate and test

Integrate the capabilities (sub-system) you just built and tested into the whole system

Test how the new set of capabilities work in the context of the whole system

8



Record and reflect

Record results

Discuss and reflect on outcomes

Decide on changes for the next iteration

9



Integrating Hardware and Software

The Engineering Process is an iterative one– “Iterative: involving repetition”

– It takes several iterations to get something right

– Don’t expect to get it right the very first time!

– Design first, then prototype with basic materials (card board etc.) then prototype with more sturdy material (more realistic prototype)

10



PART 2: GOALS AND CAPABILITIES

11



Goals drive everything

In this module you will learn the relationship between hardware and software and goals

Hardware has features and mechanisms

Software provides capabilities

Goals are things you want to achieve

Often these goals rely on more detailed goals being achieved first

To achieve any goal, you need a capability

12



Start with the Goal you want to achieve and see what other sub-goals need to be achieved first

Goals– What are we trying to do? What objective(s) do we want to achieve?

Sub-goals

Sub-goal2

Sub-goal1

Goal

Goal: to get this gear moving

In order to do that, we need to get this gear to work first

And before all else, we need to get this one going!

13



Capabilities and mechanisms allow you to achieve goals

Score on low goal

Trigger Ball Chute

Collect Balls

Use a “bush” to collect in

storage

Bush + 2 NXT Motors +

Spiral Zip Ties

Lower a ramp to let balls in

Ramp with Servo

Deploy Balls Raise Storage Area

Two Level Storage with

Servo

Goal

Capabilities Capabilities use mechanisms

14



Mechanisms enable capabilities to achieve goals

Goals Capabilities Features / Mechanisms

15





Robotics Engineering Process

<see Phase I slides>>

17



Principles

Start with Goals and Objectives; tie everything to goals– Why?

Decrease waste, time, false starts, changes, human discord

Design Before you Build or Code

Prototype/Test before you Build

Decide on an incremental set of functionality

Build up; get it working first, then enhance

Prototype with the simplest material first; prove it, then pay for the better material

As you approach the end game, freeze the changes

If it ain’t broke, don’t fix it!

Study scenarios and contingencies

Do parallel design and prototyping

Test and pick the best (the one that most easily gets you to your goal)

18



What are SCI principles related to the principles of robotic design?

List 3

19



Goal-Oriented

Goals

Capabilities– What can the robot do?

– Usually tied to software controlled set of mechanisms working together

Features/Mechanisms– Hardware

Strategies– How you use the mechanisms and capabilities to achieve the goals

20



Context and Sensors

Sensing the environment gives you context– Programming Sensors give you contextual information

Building a robot is mechanics and only becomes robotics when you imbibe it with intelligence

Intelligence comes from awareness, which comes from using sensors

21



Don’t over-automate

“Leave the driving to us” – greyhound principle

– Partially delegate driving to the robot itself, until it grows up

I.e., you add enough guard code to disallow the robot to hurt itself



Modified Asimov Laws of Robotics

23



Isaac Asimov's "Three Laws of Robotics"

A robot may not injure a human being or, through inaction, allow a human being to come to harm.

A robot must obey orders given it by human beings except where such orders would conflict with the First Law.

A robot must protect its own existence as long as such protection does not conflict with the First or Second Law.

24



Rewrite the laws to fit our competition



The Practical Functions of a DesignPart 4

26



A design's practical functions can include:

movement How will the robot move within its environment? If it were put in a different environment, would it still be able to move within this new space?

manipulation How will the robot move or manipulate other objects within its environment? Can a single robot move or manipulate more than one kind of object?

energy How is the robot powered? Can it have more than one energy source?

intelligence How does the robot "think?" What does it mean to say that a robot "thinks?"

sensing How will my robot "know" or figure out what's in its environment? If it were put in a different environment, would it be able to figure out this new environment

IBM Service Oriented Modeling and Architecture TM (SOMA) Licensed Materials - Property of IBM © Copyright IBM Corporation 2009. All rights reserved. Introduction.

Documents

engineering process

system slide

capabilities goals

capabilities form slide

system test

set of capabilities

iteration slide

realistic prototype