15-Sep-2021

Attendees

  1. @Thomas Watteyne

  2. @Kristofer PISTER

  3. @Felipe Campos

  4. @Razanne Abu-Aisheh

  5. @Said Alvarado-Marin

  6. @Alex Moreno

  7. @David Burnett

  8. @José Astorga Tobar

  9. @Razanne Abu-Aisheh

  10. @Alfonso Cortés

  11. @Andrew Fearing

Minutes

  • @Thomas Watteyne

  • @Fil Maksimovic submitted his paper on lighthouse localization to https://www.icra2022.org/

  • @Razanne Abu-Aisheh

    • Atlas works with pister-hack

    • Added Concurrent transmisison and dynamic relay placement algorithm

    • Algorithm works well but is very slow. Am working on re-architecting and optimising to make it faster.

  • @Felipe Campos @José Astorga Tobar

    • All DotBot Python package code consolidated in https://github.com/DotBots/PyDotBot

    • Implemented HTTP Orchestrator (single gateway) to control DotBot movement remotely

      • Goal: As long as Xbox controller, iPhone, or some centralized controller (e.g. Atlas) conforms to API specification, they can control a single or a swarm of DotBots via the Internet or localhost

      • MQTT skeleton in place to allow for several Gateways through the public HiveMQ broker

      • Focusing on updating Gateway and DotBot firmware

        • Passing calibration parameters & computing positions on-board DotBot (1. angle 2. x/y given 1 ceiling LH & height 3. x/y given 2 LH & homography + scale)

  • @Fil Maksimovic @Alfonso Cortés

    • Starting to work on getting a RISC-V core running on a TinyFPGA

      • Goal: flexible accelerators and peripherals

  • @Said Alvarado-Marin

    • Worked with Fil on submitting the light house localization paper.

      • Made some augmented reality plots in Blender for the paper.

    • Debugging Lighthouse glitch in DotBot.

      • Confirmed the existence of a software bug in the DotBot firmware. Bluetooth stops transmitting randomly.

  • @Mina Rady

    • Sending out PhD Manuscript today “Agile Multi-PHY Wireless Networking”.

    • Next: building a demo for a 100 km Mesh.

  • @Alex Moreno

    • boards

      • 2 Sulu variants: with regulator or LDO

      • panelized fabrication

      • $350: pannel of 48 motes, expect 5 not to work, ordered 3 panels to have 129 boards

      • ordered a stencil as well

      • order placed

    • die testing

      • looking for what is going on

      • we shoud work on nRF+SCuM test programmer

      • 9 pins

    • will be discussing with @Fil Maksimovic about VDDD

  • @David Burnett

    • Has the program counter been inspected to see which instruction a vddtap chip stalls on? We should be able to use the digital scanchain to grab the program counter without the CPU operating.