Skip to content

Meeting minutes October 2013

Jump to bottom
Jeremy Steward edited this page Nov 1, 2013 · 1 revision

23/10/2013

Attendees: SL, JS, BT, HN, AC, KA

Discussion with SL

  • Update on current plan
  • Focus on one thing you can do very well and a lot of people want (like Kickstarter)
  • Think about architecture/design
  1. Start with specifications:
  • -> Functional: "what do you want to achieve?" (e.g. we only support Green Giant cans)
  • -> Non-functional: security, privacy, availability, accuracy
  1. Think of a design
  • -> Fits specs because of...
  • Think about what analytics you can get, e.g. heat map of availabilities
  • Also think about how to put it in GIS software (mini discussion about wells and reservoirs)
  • What value can you create?
  • examples: shelf availability, people movement or path
  • Look up Ninja blocks as something we could use/get ideas from
  • (RE: Server stuff...)
  • Each application - manages its own accounts
  • Each user - can Create Read Update Delete (CRUD)
  • Super user - can CRUD everything
  • Presentation - App - Data
  • e.g. User on web page - login access (logic) - database
  • For those of us who aren't web savvy, we can go to Code School

Group-only discussion

  • Functional specs: shelf facing (i.e. is it faced or not?), motion of people
  • Update with plain text files
  • Next meeting: 1-2pm, Friday, 1 Nov
  • To do by this meeting: outline of our own sections

Lit review sections

Intro (AC, KA)

  • Purpose: problem, relevance
  • Background: current state-of-the-art, shelves, sensors, RFID, sensor server, etc...

Hardware (HN)

  • Functional specs: what have others done? (refer to background probably), what do we need
  • Non-functional specs

Software (JS)

  • Functional specs: interface, what can it do
  • Non-functional specs

Methods (BT)

  • What will we use?
  • What will we do?

Conclusion!

16/10/2013

Attendees: JS, BT, HN, AC, KA

Timeline update

  • Goal for this semester: finished prototype, possibly interface with server database
  • Reminders: lit review (Nov 18), technical deliverable presentation (as early as Nov 22...)
  • Meeting w SL some time next week (M-W)
  • Find references for your section of lit review by next week (at least five)

Group member updates & group discussion

  • Ideally want to use something cheap so that we can have a large quantity of sensors
  • Application can extend to seller side, so information would be provided to both buyers and sellers
  • e.g. two types of "shelves" (buyer and seller) so that sellers can track stock, see what sells; buyers know what to buy and where they can buy it
  • Weight sensors might be too expensive for our purposes (Wii balance board is likely the most feasible but is still too much $$)
  • Existing "smart shelf" applications, e.g. detection when an object is removed from the shelf
  • Website with list of OGC standards
  • Existing open-source interface for multiple IoT devices called OpenRemote
  • Kinect application looking at demographics and marketing at cashier shelf
  • Decided on creating TWO prototypes, one for consumers and one for sellers

Literature review

  • Have research done by next meeting (Oct 21 at the earliest)
  • General topics covered:
  • Decided application (including both "shelf" side and "interface" side)
  • Past applications (again, including both sides)
  • OGC standards
  • Data model setup (server); decentralized vs. centralized implementation
  • Internet of Things
  • Hardware options: physical (shelves, boxes), "technological" (sensors, microcontrollers)
  • Software options: e.g. C/C++, python, etc.
  • Website framework (maybe this fits somewhere else?)
  • Future challenges

See updated headings on literature review wiki

2/10/2013

Attendees: JS, BT, HN, AC, KA, SL

Discussion of possible applications

  • Fitness tracking (may be too complicated; also a lot of commercial ideas right now)
  • Parking space availability system (some existing, e.g. in SF, in YYC)
  • Snow depth (at present, humans as sensors seem to be a good option)
  • Pet monitor (up dog! - could be expanded to other options, e.g. rice, detergent, toilet paper...)
  • "Smart storage" (expansion of up dog)
  • Water quantity, e.g. flow rate, depth, etc.
  • Library books (or any "on/off" application with a magnetic sensor)

Things to think about

  • Application and Feasibility
  • Focus on low quality, high quantity (gives geospatial component)
  • Either think of applications and find sensors, or look at sensors and think of applications
  • Make a video (a la Kickstarter)

Further expansion on "smart storage" idea

  • Need a better name
  • Front end could have ads (e.g. Wal-Mart has a sale on TP)
  • Back end would look at analytics
  • Could have different sizes (e.g. small size for coffee beans, large size for TP)
  • Use wifi to hook up to home internet

Websites for future use:

Things to do:

  • Literature review: OGC Standard, similar applications already done (sensors used, systems, etc.), maybe website or app functionality?
  • Make a literature page on wiki
  • Put milestones on GitHub

Clone this wiki locally