Skip to end of metadata
Go to start of metadata

You are viewing an old version of this page. View the current version.

Compare with Current View Page History

Version 1 Next »

Object Avoidance Milestones:

  1. Bootcamp and onboarding

  2. LIDAR working - able to read outputs

  3. Get drone to stop on autonomous mission if obstacle is detected and return if obstacle is no longer detected

  4. Find path deflection algorithm in auto mode

  5. Implement path deflection algorithm in auto mode

  6. Get drone to stop on manual mode if obstacle is detected and return

  7. Path deflection in manual mode (not just stop, but stop the drone from crashing as pilot moves around the obstacle)

Weekly Milestones

  • May 6, Week 1 Milestone 1: Bootcamp

  • May 13, Week 2 Milestone 2: Test LIDAR and start obstacle avoidance repo by reading up on multiprocessing

    • Similar architecture to airside

    • Modules will be continuously integrated into main as they are created instead

    • Planning on having the co-op be the only person working on this repo, shouldn’t block other people

  • May 20, Week 3 (note, comp week): Flight interface class + workerGood starter task, architecture will be very similar to airside’s flight interface worker

    • Gets telemetry data and constantly updates the current state of the drone

  • May 27, Week 4: Detection class + worker, Decision class + worker

    • Detection (doesn’t have to have a class, depending on how the lidar works)

      • Worker constantly scan with the lidar and creates data struct

      • Passes data struct into output queue

    • Decision

      • Takes LIDAR data and decides whether or not the drone should stop

      • If it should, passes command to output queue

      • If stopped, decide if it should go back to the mission

  • Jun 3, Week 5 Milestone 3: Flight Interface input queue and testing

    • Add input queue to handle Detection Worker output

    • Reads from queue and calls upload command to stop the drone, or return the drone to the mission

  • Jun 10, Week 6: Research obstacle deflection algorithms, potential ideas include:

    • Bendy Ruler

    • Vector Field Histogram

    • Artificial Potential Fields

  • Jun 17, Week 7: Research continued?

  • Jun 24 Week 8 Milestone 4 (note, midterm evaluation): Decide on algorithm and start to implement it

    • Depending on the algorithm the architecture will be pretty different

    • At this point, halfway through the term, the co-op has done several tasks and should have an idea of how we do system design

    • I want them to start designing the architecture themself and the milestones become less structured

  • Jul 1 Week 9: Continue working on implementing path deflection in auto mode + flight testing?

  • Jul 8 Week 10 Milestone 5: Successfully tested path deflection in auto mode

  • Jul 15 Week 11: Go back to simple stopping, but in manual mode

    • Go into loiter mode if we’re going to crash

    • Based on velocity, direction, and detected objects

  • Jul 22 Week 12 Milestone 6: Successfully test stopping in manual mode

  • Jul 29 Week 13 (note, exams start: Research assistive obstacle avoidance methods, e.g.:

    • Proportional

    • Pure pursuit

    • Reactive

  • Aug 5 Week 14: Implement algorithm

  • Aug 12 Week 15: Implementation continued, Flight tests

  • Aug 19 Week 16 Milestone 7: Flight tests, offboarding

  • No labels