Date: 05/31/2020

• Galois presentation notes
  • Demos?
    • Switch between controllers
  • Show code
    • Kind 2 modifications
    • Verification attempts
    • FLC build
• Google docs
  • Report done in there then moved to LaTex
• Practice presentation Wed/Thursday/Early Friday
• Topics
  • Artem - Rust/Hardware
  • Ignacio - FLC
  • Andrew - Verification
  • Yuqi - Background & Previous team's robot

Date: 05/27/2020

Agenda

New Business

• When to do final presentation for Galois?
  • June 11th 1:00 PM
  • Find out length of presentation & if it will be virtual
• Rehearse for class presentation in one week

What has everyone done this past week?

• Artem
  • Testing
• Ignacio
  • Final controller stuff
• Yuqi
  • Wrote up background for report
• Andrew
  • Did some last ditch verification attempts
    • Problem is system is open ended and possibilities balloon as the count increases

What is everyone going to do from here?

• Artem
  • Test plan
    • C PID vs our PID
    • Other tests that are good to have
• Ignacio
  • Start writing up FL stuff for report
• Yuqi
  • Background for report
• Andrew
  • Will start slides & put link in chat

Any blocks?

• Artem
• Ignacio
• Yuqi
• Andrew

Date: 05/21/2020

Agenda

New Business

• Last technical meeting
• Artem added an integral factor to the FLC during the meeting which resulted in a much smoother motion in the robot

What has everyone done this past week?

• Artem
  • Wrote FLC in Lustre
    • Code compiled to Rust fine
    • No integral however, might be causing the oscillation
    • Actual, now working better with integral added
• Ignacio
  • Met with Matt Clark
  • Working on FLC
    • Oscillates in simulation more than hoping for
• Yuqi
  • Report stuff
    • Mathematical background
• Andrew
  • Worked on verification

What is everyone going to do from here?

• Artem
  • Testing
• Ignacio
  • FLC tweaking
• Yuqi
  • More report writing
• Andrew
  • Report writing

Any blocks?

• Artem
• Ignacio
• Yuqi
• Andrew

Date: 05/13/2020

Agenda

New Business

• Artem figured out how to eliminate parts of code that will break generated Rust in embedded

What has everyone done this past week?

• Artem
  • Figured out how to get rid of parts of generated code that won't work
  • Put generated Rust on physical hardware
• Ignacio
  • Midterms
• Yuqi
• Andrew
  • No progress

What is everyone going to do from here?

• Artem
  • Continue on making sure generated Rust is working
  • Clean up code & test
• Ignacio
  • Fuzzy logic completed
  • Work on just balance, not turning
• Yuqi
• Andrew
  • Verify working PID controller
  • More report

Any blocks?

• Artem
• Ignacio
• Yuqi
• Andrew

Date: 05/07/2020

Agenda

New Business

• Record or live presentation?
  • Live
• What time to present?
  • Friday June 5th - 1:00 to 1:15 PM

What has everyone done this past week?

• Artem
  • Converted all PID Lustre generated Rust code to embedded Rust
    • Code worked as well as PID written in Rust directly
  • Working on modifying Kind 2 to generate embedded Rust
    • Some issues with building Kind 2/ Ocaml
• Ignacio
  • Met with Matt Clark re:code provided for FLC
    • How to put into matlab
    • How to account for saturation of motors
    • Sounds like getting Rust from Matt Clark is possible and not too hard
• Yuqi
  • Busy with reports and stuff from other classes
  • Final report background
• Andrew
  • Worked on property checking PID - still not confident that I'm doing it right

What is everyone going to do from here?

• Artem
  • Continue working on modifying Kind 2
• Ignacio
  • Midterm coming up
  • Put FLC into Matlab
  • Meet with Matt Clark to continue figuring out code - Lustre specifically
• Yuqi
  • Install Kind 2 and work on that
• Andrew
  • Begin total outline of report/slides

Any blocks?

• Artem
• Ignacio
  • Time constraint due to Midterm
• Yuqi
  • Sharing computer with roommate
• Andrew

Date: 04/28/2020

Agenda

New Business

• Second robot is built and working with some minor issues
  • Set point is very sensitive, changes daily

What has everyone done this past week?

• Artem
  • Picked up parts
  • Built robot
  • Got Rust Library PID working
  • Modified Lustre generated Rust so that it compiles, very close to working with embedded
• Ignacio
  • Started looking into FL Matlab code that Matt Clark sent
• Yuqi
• Andrew
  • Administration/emails

What is everyone going to do from here?

• Artem
  • Continue working on generated Rust
• Ignacio
  • Continue looking into the code, Matt Clark is MATLAB FL Controller master
• Yuqi
  • Install Kind 2
  • Look into property checking - help support since Andrew is not able to contribute as much
  • Begin final report
• Andrew
  • Also property checking, just haven't had time to spend on it

Any blocks?

• Artem
  • Questions about having to rebuild Kind 2 every time the generator code is modified- will make an Issue for it
• Ignacio
  • Nothing major, will talk to Matt Clark if any questions
• Yuqi
• Andrew
  • Still no childcare

Date: 04/17/2020

Agenda

New Business

• Artem will meet with Chris Clark to pick up the parts

What has everyone done this past week?

• Artem
  • Trying to build Kind 2
  • Tested boards PWM output to confirm they were equivalent
    • Boards were basically equivalent, frequencies within 1%, duty cycle within 1%
  • Found that MKR 1010 was hanging up when trying to do serial data transfer
    • Means TCP/Webserver data transfer is crucial
• Ignacio
  • Bogged down with other classes/midterms/labs
  • Now free to focus on Matlab model for FL controller that Matt Clark sent
• Yuqi
• Andrew
  • Administration stuff - emails about parts

What is everyone going to do from here?

• Artem
  • Going to test serial data transfer - send constant to serial data
  • Trying to build Kind 2 still
  • Possibly meet with Chris Clark to pickup parts
• Ignacio
  • Work out Matlab model that Matt Clark sent for FL controller
• Yuqi
• Andrew
  • Artem Lustre - test properties
  • Get in touch with Ethan

Any blocks?

• Artem
• Ignacio
  • Minor computer issue
• Yuqi
• Andrew

Date: 04/09/2020

Agenda

New Business

• Do we want to make slides?
  • No
  • Just screen share
• Andrew - discuss PDS
• Artem & Ignacio - discuss test plan

What has everyone done this past week?

• Artem
  • Got structs to work in Rust - hooray
    • Should make translating lib from Lustre generation will be easier
  • Wrote PID in Lustre
    • Wanted to get as close as possible to controller in Rust
  • Worked on test plan
• Ignacio
  • Test plan creation/refinement
  • Simulink working with physical constraints of motors
• Yuqi
  • PWM test plan work
    • Researched how to write a program to test PWM output
• Andrew
  • Worked on previous team's PID Lustre script to check properties

What is everyone going to do from here?

• Artem
  • Test PID libraries to confirm that they are both working the same
  • Work on testing the two boards
    • Want to set up next team
• Ignacio
  • General simulation refinement
  • Other classes need to take precedence
• Yuqi
  • Email professor Faust to find out what is going on with the hardware
• Andrew
  • Check with Ethan to confirm property verification
  • Check properties on PID from Artem's Lustre

Any blocks?

• Artem
  • Need's battery charger
• Ignacio
  • Lots of stuff due in other classes/midterms
  • Waiting on Matt Clark for FL stuff
• Yuqi
• Andrew

Date: 04/03/2020

Post Class Check-in Meeting

Agenda

New Business

• New PDS
  • Remove physical measurement requirements
  • Figure out details and then schedule with Galois
    • Try to meet with Galois on Tuesday
    • Let them know about class requirement for revised PDS
• Test plan
  • Focus on old robot
  • Test IMU to webserver data capture
    • Is it capturing tilt & yaw correctly?
    • Is there significant data loss?

What is everyone going to do from here?

• Artem
• Ignacio
• Yuqi
• Andrew
  • Email Galois
  • Revised PDS

Date: 03/31/2020

Agenda

New Business

• Tuesday - mandatory meeting, Friday - depending on tasks

What has everyone done this past week?

• Artem
  • Sanity break
• Ignacio
  • Researched physical limitations of motors
  • Found constraints - max torque (stall torque - 0.1 NM)
  • Tested simulation with 70% of max torque, FL controller works okay
  • Contacted Matt Clark - he has a Lustre implementation of FL controller - possibly able to help us convert ours to his form, waiting to hear back
• Yuqi
  • Working on test plan
• Andrew
  • Childcare

What will everyone do this upcoming weeks?

• Artem
  • PID Lustre
  • Test Plan review
• Ignacio
  • Improve FL controller
  • Attempt to integrate pitch and yaw controllers
  • Continue talking with Matt Clark about implementation and Lustre conversion
  • Test Plan review
• Yuqi
  • Test Plan review
• Andrew
  • Verification of specific properties in Lustre
  • Test Plan review

Any blocks?

• Ethan

---

---

Date: 03/24/2020

Agenda

New Business

• Special projects room is closed until further notice - means that we can't get parts for new robots
• We need a charger for the old robot
• Sending status update

What has everyone done this past week?

• Artem
  • Researched Simulink to Lustre conversion
    • There is a free tool (S2L) but uses Matlab 2013 - possible solution
  • Trying to write Lustre PID for verification - need more research into Lustre
• Ignacio
  • Wrapped up Fuzzy Logic simulation
• Yuqi
  • Recovering but feeling better
• Andrew
  • Sent some emails
  • Lack of daycare means almost no time during the day to work

What will everyone do this upcoming weeks?

• Artem
  • Implementing Lustre PID
• Ignacio
  • Looking into practicality of implementing the FL controller - are motors capable of providing necessary torque, etc.
• Yuqi
  • Fleshing out test plan - closure of lab means that we will probably only be comparing old robot physical with simulation
• Andrew
  • Look into exactly how to specify the properties that we want to check in Kind 2

Any blocks?

• Coronavirus
• Lab is closed

---

---

Date: 03/19/2020

Agenda

New Business

• Andrew's kid's daycare is closed, amount of time to work on Capstone is reduced
• Yuqi is still healing from stomach
• Went over remaining tasks and what is remaining
  • Good amount done but quite a bit left
• Going to try to build robot by March 29th
• Potentially build robot on Wed - 25th

What has everyone done this past week?

• Artem
  • Lustre work
• Ignacio
  • Fuzzy logic, but also just wrapping up finals/papers
• Yuqi
  • Test plan/health recovery
• Andrew
  • Stalled with childcare

What will everyone do this upcoming weeks?

• Artem
  • Compile Rust from Lustre of PID
• Ignacio
  • Simulink to Lustre
• Yuqi
  • Finalize test plan
• Andrew
  • Pick up parts and characterize

Any blocks?

• Coronavirus

---

---

Date: 03/14/2020

Agenda

New Business

• Can we even build new robot?
  • Checking in with Ethan and Faust about getting parts
• Feeling like we are running out of time
  • We will look at how many actual days left
• Yuqi not attending meetings
• Virus has suspended classes/in-person meeting
• Unclear if we will be able to get parts

What has everyone done this past week?

• Artem
  • Webserver for data logging implementation completed
  • Generated Rust from plant example
• Ignacio
  • Bogged down with finals
• Yuqi
• Andrew
  • Bogged down with finals

What will everyone do this upcoming weeks?

• Everyone bogged down with finals
• Artem
  • Kind 2 support/research
• Ignacio
  • Lock down Fuzzy logic - with psuedocode
  • Matlab code ready to translate to embedded
• Yuqi
• Andrew
  • Email Faust/Ethan
  • Lustre
  • Characterize parts before building
  • Try to build robot, pending parts

Any blocks?

• Artem
  • Finals not over
• Ignacio
  • Finals not over
• Yuqi
• Andrew
  • Waiting for parts still

---

---

Date: 03/07/2020

Agenda

New Business

• Dead week slowdown

What has everyone done this past week?

• Artem
  • Webserver implementation
    • TCP in Rust
    • Reliability achieved - hours of consistent testing
• Ignacio
  • Controller testing
  • Fuzzy logic & other controller tweaking
  • Fuzzy logic is working well
• Yuqi
• Andrew
  • Emails out to Ethan & Tymerski
  • Reading Lustre Manual
  • Still working on implementing Lustre code for Arduino PID

What will everyone do this upcoming week?

• Everyone bogged down with finals
• Artem
• Ignacio
  • Research Simulink to Lustre
• Yuqi
• Andrew
  • Get parts for robot

Any blocks?

• Artem
• Ignacio
• Yuqi
• Andrew
  • Waiting for parts

---

---

Date: 02/29/2020

Agenda

New Business

• Going to focus on building an updated robot as that will allow us to test new controller/finish first issue/have a tangible piece of evidence that we did something
• Meet in person? After we get parts

What has everyone done this past week?

• Artem
  • Webserver
  • Troubleshooting hardware (IMU specifically)
  • Trying to get it to be high assurance
• Ignacio
  • Fuzzy logic PD controller for balancing
  • More research into designing a better fuzzy logic controller
  • Hoping to implement an adaptive fuzzy logic controller and researching that
• Yuqi
  • BOM for new robot
  • Rough test plan creation
• Andrew
  • Kind 2 testing
  • Reading about Lustre

What will everyone do this upcoming week?

• Artem
  • Continue work on webserver/getting data out of robot
  • Help with new robot
• Ignacio
  • Continue fuzzy logic controller research/refinement
  • Book contains adaptive controller information/sliding mode/neural network, so will look into that
  • Help with new robot
• Yuqi
  • Continue test plan creation
  • Build robot
  • Upload BOM
• Andrew
  • Implement Arduino PID in Lustre
  • Email about grading
  • Email about new robot parts

Any blocks?

• Artem
  • Processor might be too slow? 0.5 speed of the previous processor
  • Need battery charger
• Ignacio
• Yuqi
  • Measurement of RPM
• Andrew

---

---

Date: 02/21/2020

Agenda

New Business

• Nothing

What has everyone done this past week?

• Artem
  • Fixed PWM
  • Commented out BT
  • Compiled PID library in Rust
  • Replaced ESP32 board with MKR1010 board on Robot
  • Ran PID Rust Library on Robot & confirmed that it works
• Ignacio
  • Confirmed with Ethan that nominal model is good to go
  • Researching runtime verification
  • Researching different types of controllers (fuzzy logic?)
• Yuqi
  • Test plan creation
• Andrew
  • Got kind2 working
  • Emailed Tom Schubert
  • Read formal verification book

What will everyone do this upcoming week?

• Artem
  • Webserver implementation research and development
  • Document first issue completion
  • Possibly fix wobble (not first or second priority)
• Ignacio
  • Choose controller
  • Implement in Simulink
• Yuqi
  • Continue test plan creation
• Andrew
  • Implement Arduino PID in Lustre
  • Compare Kind2 output to Artem's Rust PID
  • Find out what Kind2 generates for Rust - equation?

Any blocks?

• Artem
  • Resolder wiring - will check with Galois if needed
• Ignacio
• Yuqi
• Andrew

---

---

---

Date: 02/14/2020

Agenda

New Business

• Lets get to work!
• We will meet on Fridays and Tuesdays around 10:30-11:00am

What has everyone done this past week?

• Artem
  • Presentation
  • Project Schedule done
  • Typed up minutes
• Ignacio
  • Presentation
  • Project Schedule done
  • Researching sliding mode controller - good progress, has been done before so definitely doable
  • Provide State Space resources
• Yuqi
  • Presentation
• Andrew
  • Presentation
  • Expanded PDS
  • Read formal verification book

What will everyone do this upcoming week?

• Artem
  • Create plan to port code to MKR1010
  • Start porting code to MKR1010
• Ignacio
  • Work on defining a "nominal" model for the robot (Ethan may help)
  • Sliding mode controller research
• Yuqi
  • Get robot from Ethan
  • Look at improvements/plan how to test new components for model
• Andrew
  • Reach out to Tom Schubert re: Verification
  • Continue reading formal verification
  • Research state space representations

Any blocks?

• Artem
• Ignacio
• Yuqi
  • Need robot/parts
• Andrew

---

---

---

Date: 02/04/2020

Agenda

• Presentation
  • Who wants to take the lead?
    • Andrew will take lead
    • We will practice Friday at 11:40 - IEEE Lounge

What has everyone done this past week?

• Everyone met with Galois
• Artem
  • Project Schedule
• Ignacio
  • PDS revision & Project Schedule
  • Gave PDS to Tymerski - he said as long as sponsors are happy than he is happy
• Yuqi
  • Trying to get new robot stuff
• Andrew
  • PDS revision & reading Formal Verification book

What will everyone do this upcoming week?

• Artem
  • Complete project schedule
  • Track down robot
  • Investigate bluetooth code
  • Type up Galois minutes from 01/28/2020
• Ignacio
  • Complete project schedule
  • Controls research - this will be ongoing
• Yuqi
  • Track down robot
  • Look into building new robot
• Andrew
  • Expand intro to PDS for more background
  • Complete project schedule
  • Take the lead on the presentation

Any blocks?

• Artem
• Ignacio
• Yuqi
  • Waiting on Ethan for robot location - will message on Discord
• Andrew

---

Date: 01/13/2020 Agenda

What has everyone done this past week?

• Artem
  • Working through Rust book - halfway through
  • Embedded Rust book is next
  • Looked at current code, discovered bluetooth works but wifi having trouble
  • Bluetooth module needs to be rewritten for new board
• Ignacio
  • Did some modeling work/investigation
  • Reading book - Advanced Control
  • Looking at previous team's work
  • Going to reach out to Ethan/Galois to help with non-linear controller
• Yuqi
  • Block diagrams for PDS/reference
  • Read last report/code
  • Creating diagram for PID
• Andrew
  • Did a little PDS work - realized we each need to do the background for the part of the project we are going to be responsible
  • Researched formal verification
  • Practiced Lustre

What still have to be completed?

• PDS Notes
  • What has been done, and how much is left to do?
    • Bare layout, need to have each person fill in their background.
    • Andrew will take all text in Google Doc and move to Overleaf
  • What help is needed with this task?
    • Everyone will do their own section
  • Where can we view a copy of the completed work?
    • It will be located on Overleaf until it's deemed "acceptable" and then will be on the Repo
  • How much (if any) can be taken from last year's PDS?
    • Guessing very little since they focused on the physical aspect (building the robot).
• Project Schedule notes
  • What has been done, and how much is left to do?
    • Research has been done, should be added
    • Most of the schedule to be done, will be easier after PDS is complete
  • What help is needed with this task?
    • PDS needs to be completed, Artem will put first month-ish of stuff in (get new board working, move PID to Rust, build new robot, etc.)

Questions:

• Do we want to send a copy of the early PDS to Galois?
  • Yes, will send (hopefully) on the 21st
• Current issue: C + Rust (PID controller code) working on a robot:
  • How much time should this take?
    • 1 month?
  • How many people to allocate to this task?
    • 2-3
  • Does anyone want to branch off and focus on something else during this time?
    • Ignacio will be doing controller research
• Dress code for presentation?
  • No

Question for Galois:

• Is non-linear controller feasible?
• What about sliding mode control?
• Code need to be implemented in Lustre?
• Can Ethan come to meeting?

Issues:

• Bluetooth module on MKR 1010 is different from bluetooth module on previous project
  • Bluetooth code will likely need a rewrite before compile.
• Need to do a lot of research to get up to point of understanding formal verification.

Reminders: 28 January, PDS meeting at 10:30am

---

Date: 01/07/2020

Agenda

• When do we want to complete the PDS and project schedule.
• How are we going to divide the work for the PDS and schedule.
  • Understand general structure of previous teams Otis_arduino code and translate it at a high level.
  • Create a high level diagram of how to translate the code to the Arduino MKR WIFI 1010.
  • Understand how to translate the PID control to a Rust library.
  • A high level model of a nominal self balancing robot and the actual robot.
  • The high level structure of the nonlinear controller(actual structure and form of verification)
• How are we going to divide the work for the project in general.
  • Project Management
  • Programming
  • Controls
  • Modeling
  • etc.

Notes

• Talked about priorities, sounds like right now we want to focus on getting PDS and Project schedule completed, with research happening as well.

Capstone Stuff

• PDS - Andrew
  • Reach out to Faust for detail level
• Project Schedule Prelim - Artem

General Project Areas

• Controls - Ignacio
• Rust - Artem
• Verification - Andrew
  • Do we need physical robot to do verification?
• New robot - Yuqi
• Read last year's report - Everyone except Ignacio

Questions

• Is building the second robot a stretch goal?
• What is the status of the PID control in otis? Does it need to be converted to Rust? What is the Rust PID controller library mentioned in the report?
• Is there a MATLAB model? Whats the status?

---