Sprint 2

Introduction

This sprint builds on top of the work of sprint 1 and aims to integrate the last component of the system: the cargorobot.

Requirement Analysis

The requirement analysis has already been specified in the sprint 0, this particular sprint does not require anything else to be specified in this section.

Problem Analysis

Using the previous requirement analysis as source, we will briefly summarize here the behaviour of the cargorobot actor:

1. It moves from the HOME to the IOPORT
2. It takes the container from IOPORT to the slot-5
3. Waits for the marking to be done
4. Moves the container from the slot-5 to the assigned slot
5. It goes back to the HOME

As described in the sprint 0 we have a few implementations of an already existing cargorobot to choose from:

• VirtualRobot26
• RobotObj26
• RobotService26
• RobotSmart26

The implementation that matches more closely our needs is RobotSmart26 because:

• We need to move the robot between places that are not always fixed and known beforehand, so a pathfinding service is needed in our use case.
• The system need to stay reactive during the robot’s movement, so an asynchronous implementation is needed.
• The robot uses a map system that closely resembles our hold implementation.

The code of RobotSmart26 is available on the repository, since our project shouldn’t modify the code, it will be compiled as is and the rest of the project will interact only by the qak messages:

Request buildPlan : buildPlan(PX,PY,TX,TY)
Reply buildPlanDone : buildPlanDone( PLAN ) for buildPlan

Request moverobot : moverobot(TARGETX, TARGETY,STEPTIME)
Reply moverobotdone : moverobotok(ARG) for moverobot
Reply moverobotfailed :  moverobotfailed(PLANDONE, PLANTODO) for moverobot

Dispatch noplan : noplan(X)
Dispatch setplanbuildelay : value(V)

Request  doplan : doplan(PLAN, STEPTIME)
Reply doplandone : doplandone(ARG) for doplan
Reply doplanfailed : doplanfailed(PLANTODO) for doplan

Request step : step(TIME)
Reply stepdone : stepdone(V) for step
Reply stepfailed : stepfailed(DURATION, CAUSE) for step

Dispatch move : move(M)
Dispatch setrobotstate : setpos(X,Y,D)

Request setdirection : dir(D)
Reply setdirectiondone : pos(PX,PY) for setdirection

Request tuneAtHome : tuneAtHome(X)
Reply tuneDone : tuneDone(X) for tuneAtHome

Request getrobotstate : getrobotstate(ARG)
Reply robotstate : robotstate(POS,DIR) for getrobotstate

Test Plans

By the end of the sprint every component of the system has implemented in its entirety, so the user can fully test each part and see if it meets the reqirements. As in the sprint 1 the tests are end-to-end and no unit tests are required. To see if the system works, a virtual environment containing a simulation of the robot and the hold is available as a web page.

Deployment

The system will need two more containers:

• The cargorobot subsystem
• A virtual environment to test the system

As said before, because the cargorobot’s code should not be touched we decided to distribute the precompiled code in a single tar file. The system will also need a couple files to load the hold map and some initialization parameters

FROM eclipse-temurin:17.0.5_8-jre-focal AS builder

ADD ./external/robotsmart26/robotsmart26-1.0.tar /

WORKDIR /robotsmart26-1.0/bin

COPY ./*.pl ./
COPY ./external/robotsmart26/basicrobotParams.json ./
COPY ./*.bin ./
COPY ./external/robotsmart26/tf25map.txt ./

CMD ["bash", "robotsmart26"]

The wenv instead will simply be pulled from the docker registry

# Other services from sprint 1
# ...
wenv:
    container_name: wenv
    image: docker.io/natbodocker/virtualrobotdisi26:1.0
    ports:
      - 8090:8090 
      - 8091:8091/tcp
      - 8091:8091/udp
    restart: unless-stopped

  robotoutgui25:
    container_name: robotoutgui25
    image: docker.io/natbodocker/robotoutgui25:1.0
    ports:
     - 8085:8085/tcp
    restart: unless-stopped

  robotsmart26:
    container_name: robotsmart26
    build:
      dockerfile: ./docker/Dockerfile.robotsmart26
    ports:
      - "8020:8020/tcp"
      - "8020:8020/udp"
    environment:
      - VIRTUAL_ENV=wenv
      - MQTTBROKER=mosquitto
    depends_on:
      - wenv
    restart: unless-stopped