shortcuts:
lectures
homeworks
lab notebook
search:
Path Planning
(Tue Oct 22, lect 15)
|
previous
|
next
|
slides
|
Look a little deeper at how its done
Logistics
Delay “hello robot” to Sunday, ok?
Please note how to submit team homework
Quiz is coming on March 29
Working In Teams
In all my courses I ask students to work on projects in teams.
I have seen them succeed and have seen them struggle.
These notes are based on this hard won experience with student run teams.
What kinds of ideas or approaches have you used (or been used on you) that you think are good?
Lets hear some horror stories?
Establishing Norms / Social Contract
NB
You're not trying to become best friends with your teammates. You're trying to get something done!
A key challenge is to establish the “norms” or the “culture” of your team.
What I mean by that is the unspoken or spoken etiquette or understandings.
For example, how often you meet, how much you care about being punctual, how you give each other feedback, how you deal with conflict.
Each team, even of just four people, might handle things a little differently.
A college course, which runs only 13 weeks, this there’s not a whole lot of time for these things to happen organically.
This means that everyone needs to pitch in a little extra to it a fun and useful experience for everyone.
Time to meet
One of the big challenges is to find time to meet, so try to be flexible.
As a team, choose a fixed time and place to meet every week. Be flexible.
Your first meeting is very important.
Personal Goals
Each member honestly say what your personal goals are for this course
This course is my most important, I really care and will do anything to be successful
Or, I have a heavy course load, I will do what I can but I have to balance
Or, I am perfectly happy with a “B”, don’t expect intensity from me.
If everyone knows each other’s perspective this will set expectations
Each member of the team is responsible for their own work and for the wellbeing of the team
Decisions
Talk about how you will make decisions
Democratic vote
Decide or chose a leader to have final say
Divide up the domain (front end/back end/ etc) and assign it to people.
Be sensitive to everyone’s particular constraints
Be kind to each other
My belief is that every single person on the team is doing their best
given
their own situation.
If it seems like someone is not doing their best, or somehow being counterproductive
Try to figure out their situation or ask them.
Once you understand the context things often make sense and you can leave any anger or resentment behind.
Commitment and enthusiasm
Keep each other to a high standard
Be enthusiastic and committed
If one person didn’t hold their end up this week, don’t let that discourage you or lead you to not hold up your end next week.
Path Planning
For now, just focus on a mobile robot like a car
Goal of path planning: get the robot from where it is to a destination
Beyond that there are many variations
What kind of map is present?
Are the obstacles encountered known ahead of time, and can they move?
Types of maps - Graphs
Define places where Robot can be (nodes)
Define which ones are connected (edges)
Edge information includes
What path to take
What the “cost” would be of taking that path (surface, distance, desirability)
Planning the route becomes a graph searching problem
Find the “cheapest” path between two nodes
Types of maps - Occupancy Grids
Divide the “area” into a regular grid
Indicate whether each “cell” in a grid is occupied or free
Occupancy Grids can be multi-layer
For example to indicate
access for different types of vehicles
areas that are free but discouraged
Treat each cell in the grid as a
node
with 4 or 8 edges
Thank you. Questions?
(random Image from picsum.photos)
Welcome
Autonomous Robotics
List of Lectures
Homeworks
Final Projects
Programming Assignments
Lectures
Sensing and Moving
Welcome
ROS From the top
ROS Nodes and Topics
Robot Hardware
Sensing with Lidar
Perception and Sensors
How do mobile Robots move?
ROS Programming Best Practices
Orienting
Robot Orientation
ROS Coordinates
The almighty TF package
How do Robots Localize?
Localization in practice
Navigating
Path Planning
Advanced Topics
Fiducials
Finite State Machines
Behavior Trees
Intro to computer Vision
Kalman Filters
Kalman Filters
Joints and Links
Manipulators
Projects
Project Work
Project Work
Last Class!
Course Resources
Guide for Teaching Asssistants!
References
Cribsheets
Key Papers in Robotics
Book Supplements
Cosi119a FAQ
The "Big Ideas"