Hardware session 9:45 8 May 2018

Open Ag Technology GOAT 2018
1

Summary topics discussed

  • What re people using what they’re using
  • Documenting the Hardware design processes and component selection
  • Can we build an online community: who’s going to lead that?
  • LoRa
  • HW for actual farmers not geeks having fun…
  • Drones
  • Imaging hardware

TODOS/Takeaways

  • Lack of documenting hardware design process is a problem - information isn’t readily available
  • Can we make a RIOT channel/thread on the forum maybe?
  • Maybe further down the line we could create webpage/wiki/gitpages?
  • LoRa coverage mapping experiments in Ag contexts are needed
  • Could we group purchase some expensive components for cost saving
  • If you are interested in hyper spec dev watch the forums for Greg to post
  • There’s a missing link between Ag and engineering even though lots of people are working on it - missing link due to documentation?
    – Also sometimes it’s actual language barrier (english/french named)
    – Farmhack/OpenAg/ what else can be done to encourage connections?
    - People have to want to do it
  • Look at Vinduino: tinker/hobby farmer (viticulture) building stuff
  • Anyone want to geek out on ADC/VReg etc specs - David would be keen

Sticky topics:

  • New technologies people are working in
  • The process of developing hardware

Attending, key skill/current project, interest from the session:

  • Brian Davis: Data processing, connects hardware team with end user, interested in best pracitces
  • Ned Horning: Tinker, imaging hardware
  • Gwen Jones: Works with international farmer communities, lots of mechanical open hardware experience
  • Will Szal: RegenNetwork, monitoring technologies, what hardware people are using
  • David Forester: Forum convener re HW, Tinkerer, has a wiki to share, studies spec sheets, wants to understanding component selection process
  • Patrick Oconner: HW testing and documentation, Event planning, polays with SOCs, embedded SW
  • Jane Wyngaard: Drones, EE
  • Daniel Fisher(aka Ken): Open Source HW, USDA, open hardware for new sensing and monitoring systems, arduino, data logging systems
  • Frank Rose: Software, ThingsNetwork LoRa, Arduinos
  • Samual Oslo: Farmers, Working with Farmers who are interested in HW developed in Quebec, interested in dev process/sharing process
  • Reginald Fletcher: Hyper spectral and rmote sensing/camera systems

Process of selecting components

  • The front person often in the dark when demoing gear to users about why xyz decision was made re the hardware
  • We need project metadata
  • Can we setup a monthly call/mailing list?
  • Crowd supply process: they have continuous blogging of their process and decision making often to keep their supports informed
  • Personal Wikis
  • Sphinx
  • In industry often decision is driven by convenience (which chip is in my database for schematic design/available from my supplier)

How do we disseminate HW/info/methods/…

  • Time line from solution to publication…
  • Git pages
  • SW thinks code documents itself - less easy for HW - schematics semi self documenting but requires high level of expertise to read

LoRa

- LoRa/LoRaWAN for rural areas?
- Long Range, Low data rate, Low power
- We need emperical data about range etc in forestsfarmland/....
- Ken using some for local farm work in dev
- Others using cell but many areas don't have cellular signal
- Prices: ~$10-40
- Coverage mapping of LoRa needed
- ISM bands so bands can't be bought
- Protocol has features to mitigate interference etc
- Proprietary chip
- Protocol is open
- Runs in an unregulated space
- 915MHz in US/856in EU

Imaging

  • Problem: deploy cameras in a field to watch corn growing and note when leaf curls to detect water stress - possible? - concluded yes

    • Could and/or use reflectivity/other
    • Should be do-able
    • HW is the easy bit of this project - relatively speaking

  • HyperSpec camera
    - will be very useful for predictive imaging - see things the human eye can’t
    - Single pixel camera version for low cost high res solution
    - Needs a low cost random gridding mechanism
    - Can lower your noise floor using the random mask
    - Greg is going to build one in the next year (we think) - anyone want to be involved watch the forum
    - Uses: differentiate between vegetation (alien detection), soil science (distinguish between different minerals)

    • There are new multispec sensors available for low cost (AMS sensor) ~$25

Drones

- Hyper or not for Ag specifically? - still evolving
- One place they'd definitely be useful perhaps is for predictive imaging-  Can drone see something before humans? (needs hyperspec cameras)
- no fly zones: could use kites, balloons

HW for actual farmers

  • Alot of tech geeks with cool gear but what about actual tech for Ag:
    • Alot of us are solving our own problems
    • How to match problem to engineer?
    • Some forums for facilitating connecting engineers to farmer problems: FarmHack, OpenAg,…others?
  • Tinkeres are seeing tech from say home automation/energy monitoring and choosing to apply it to Ag problems - there’s a translation barrier here
  • Farmers are probably best at looking out at the world and seeing if they can mod things for Ag
  • There’s a missing link to documentation…
1 Like