Skip to content
- Reuse whatever we can from the old code
- Obstacle drawing
- Algorithms themselves
- Showed her the progress on the PID code
- Need to get REACT working
- Maybe multiple visualizations, could separate the visualization from the code already here since it’s pretty general
- Want to make the tools accessible broadly
- If we finish bug algorithms and PID:
- Having a common UI across multiple aspect
- We have RRT and a kinematics model, but the UI is not great
- Different UI than obstacle creation, but incorporating
- Visibility graph (partial stretch) or PRM (major stretch)
- Using obstacle configuration from Bug Algs
- Focus on UI quality before any stretch goals
- Make UI generally usable so it can incorporate other algorithms later!
- As distinct as possible from the planning part
- Making the code as portable as possible
- Easily shareable
- Minimal installation
- Backend not needed
- Need to review architecture diagram, and reflect that there’s no backend on it
- Redesign the UI/UX
- Framework if we got it might be different but the actual path finding code would be useful
- Keep the copyright notice and license, modify it from there
- Can safely start a new code base, but just get what we need from the old code
- Survey some animation JS library so we can plot and design the robot, etc.
- Discussed the midterm presentation
- Harin & Sarah: platform selection and architecture diagram
- Alexandra & Zuntue: continue working on algorithms
- Ideas for PID
- Draw a line and then the robot has to follow the line. Ball on hill idea.
- Make the robot just a point rather than a differential drive.
- PID controller representation ideas
- Having the user draw a line, and robot has to follow it
- Every time line changes direction, can the robot follow
- Similar to wall following
- Making a hill/wedge, having something go up it and stop at a specific spot
- Might need gravity/acceleration
- A bit more abstract, might prefer the other idea
- A wheel that spins fast getting it to a certain RPM
- Ron really wants to have it self-contained
- Asked for more info on how his website works and how it’s integrated
- Janine will email IT about the other cs robot website about how to embed it
- We should research this a bit more too
- Divided up the work
- Harin: update website (deliverable, journal)
- Sarah: update website (project concept/about)
- Zunute: research how to implement the PID controllers
- Alexandra: research how to implement the bug algorithms
- Take a look at the previous code, ask Janine if we should keep it and build on top of it
- Current UI is hard to use
- Need to look at ethics assignment for next weekend
- Midterm presentation is in three weeks, start thinking about it
- https://www.microsoft.com/en-us/research/academic-program/give-great-research-talk/
- Helpful video on doing a research talk
- Research existing JS libraries
- We should set a goal for these two weeks, because we need to get started on it
- Implement bug algorithm after we’ve researched it
- Put all materials on website
- Can also put user stories doc under deliverables
- Need to further update journal
- Update project concept
- Show Janine the user stories and project concept
- Look at the code from the git repo from last semester
- Divided up the work
- Harin: update website (journal, side bar, team tab formatting)
- Alexandra: update website (about, links)
- Sarah: user story
- Zuntue: requirements
- First client meeting with Dr. Janine Hoelscher
- Look at existing document and video
- Make sure certain requirements have been met from last semester
- See what works and what doesn’t work from last semester’s project
- They built a lot of infrastructure, and one more complex algorithm
- They built a tool that lets you draw obstacles
- Self-contained as possible so it can be integrated into a website pretty easily
- Check for dependencies
- Everything in one folder
- Adding path planning algorithms
- Shouldn’t be too difficult to implement, a lot of code has been implemented already
- Add an easy visualization for PID controllers
- There isn’t a really good simple explanation for this already, so it’d be good to have on this
- https://robotics.cs.unc.edu/C-space/index.html
- This is what Ron has in mind
- They also tried to build kinematics models, which visually could use improvement