Manager   •   about 5 years ago

What hardware/materials should we buy and have waiting for you?

Resistors, transistors, wires, soldering stations, microcontrollers, sensors, switches, motors, raspberry pis, arduinos, beacons ....

We'll get a pile of stuff for you to use. What should we get?

  • 19 comments

  •   •   about 5 years ago

    @Shane Reiser you should not buy anything like this we have a compatible solution for this we dont need to have any microcontroller etc

  •   •   about 5 years ago

    We measured the chillers near by our home its length was 51 inches and its width was 19 inches plus its one sided square was round about 16 inches

  •   •   about 5 years ago

    @Shane Reiser
    This is a list of questions and comments from my team, iudicium:
    The hardware list you started with is great; to clarify when you said switches did you mean:
    power switches
    push-button
    limit switches
    1. What will we have to attach the hardware (e.g. push-button, limit switches, and resistors) to?
    Is something like a Soldering Board provided?

    2. What materials do we have available to secure the soldering board and sensors to the correct location in the cooler once our electronics are attached?
    Woodshop, wood, plastic sheet, drill, drill bits, zip ties, duct tape, 3d printed part?

    3. At each of the cooler locations is internet available/expected(wireless or hardline) or does the solution need to be wireless via GSM(ATT,Verizon) or CDMA(Sprint)?

    4. Do you have a database server available to store data in and run queries against?

    "None of the previous approaches achieved the scale or creation of high quality data required to make a meaningful improvement in the system's ability to predict, identify, and address out of stocks."
    Identifying out of stocks is a portion of the challenge. "

    5. Can you give an example of what a meaningful improvement to your prediction would look like?
    e.g.
    Cooler#00000000 is estimated to be sold out of Drink[N] in [I] hours.
    Cooler#00000000 is estimated to be sold out of Drink[X] in [Z] hours.
    Cooler#00000001 is estimated to be sold out of Drink[N], Drink[X] in [Y] hours.

    Or would it look something like
    e.g.
    Cooler#00000000 on tuesdays-thursday coke classic sells out while coke zero doesn't sell more than a few. It is recommended that you reduce the # of coke zero bays by *algorithmically determined value* and replace it with coke classic.
    Cooler#00000000 on saturday coke classic does not sell more than 50% while sprite, minute maid, and water all sell out. It is recommended that you reduce the # of coke classic bays by *algorithmically determined value* and replace bays 4, 5,6 with: sprite, minute maid, and water.
    Cooler#00000001 never sells more than 50% of any items on current restocking schedule. It is recommended, unless Cooler#00000036 (...nearby cooler list[n]) is being restocked.

    6. We saw other posts about power management that prompted some discussion about power management but came to the conclusion that:
    1. It didn't matter for the hackathon but is worth thinking about for scaling out to units worldwide
    2. the cooler is plugged in and long term we should be able to interface with the current power management system
    3. the cooler in refrigeration uses way more power (1000x+) than our most power hungry idea (cellular data) so power consumption isn't a concern
    Are the above assumptions correct?

    7. In Coca Cola's previous failed attempts how accurate was your data? We have discussed perfect systems(100% accurate) that don't scale due to cost or depend on other changes (modifying drink containers is out of the question, we assume) all the way through systems that are economically very reasonable for the data performance. Where is the line between "good enough" and "not good enough" for high quality data?

    8. Although not our current best approach we are quite curious as to how your cameras were placed and the data analysis done – as numerous of us are quite interested in machine learning and vision. Are the case studies available?

    9. Do you have a dataset of bottle and can sizes?

    Thank you for enabling us to further grok the intent of the competition.

    Team Leader for iudicium
    -David

  •   •   about 5 years ago

    excuse me shane reiser what is the agenda of this project
    i think that if we remove the products other than coca cola product from coke cooler that would be the highest achievement of this project

  • Manager   •   about 5 years ago

    @Abdul - the schedule is posted here - http://coolerhack.devpost.com. Scroll down.

  •   •   about 5 years ago

    @ Abdul Nasir
    They stock coolers with all of their drink products
    Find a list of soda here:
    http://www.cokesolutions.com/OurProducts/Pages/Site%20Pages/SoftDrinks.aspx
    but keep in mind it is not limited to those drinks as Coca Cola also sells water, juice, sports drinks, etc
    In the spirit of coopertition, after reading quoted portions below, is there anything else you are wondering about the challenge that I can help clarify? (We have been thinking about this for awhile now)

    "The Challenge:
    Help Coca-Cola keep coolers stocked and get people the drinks they want. Hack on drink coolers and design a scalable solution using whatever technology you think can get the job done.

    More background - i.e. the exact problem we are trying to solve?
    Out-of-Stock is the #1 Strategic Issue at Coca-Cola: This problem has resisted traditional technology solutions; from sensors on coolers to mobile devices in the hands of consumers to cameras. None of the previous approaches achieved the scale or creation of high quality data required to make a meaningful improvement in the system's ability to predict, identify, and address out of stocks.

    We need some fresh minds and outside perspective to help us solve this problem. What else you got?
    "-http://coolerhack.devpost.com.

    Team Leader for iudicium
    -David

  • Manager   •   about 5 years ago

    @David Flanigan -

    I answered some below and had to pass the rest on to the experts.

    1. We'll have two soldering irons. If you have an iron or a soldering station please bring it, too.

    2. We should have a couple 3D printers there for you to quickly print mounts. We'll get some duct tape and zip ties, too. Please bring any wood, screws, nails you need.

    3. I'll post this in #coaches and let them answer

    4. I'll post this in #coaches and let them answer

    5. I'll post this in #coaches and let them answer

    6. Those assumptions are good.

    7. That's a question I need to ask someone very high up at Coke. I'll get back to you. I'll also post in #coaches in case they have opinions.

    8. I haven't been able to get the cameras case study yet, but still trying.

    9. There are many (hundreds) of different sizes, shapes and weights of Coke products in many different package types, and they are always changing and will undoubtedly continue to change in unpredictable ways. In order for a solution to be scalable, it mustn't be calibrated for any specific set of items.

    I'm buying hardware soon. Can you send me links to the power switches, push-buttons, limit switches, etc that you need?

  •   •   about 5 years ago

    Thank you @Shane Reiser,
    9. There are many (hundreds) of different sizes, shapes and weights of Coke products in many different package types, and they are always changing and will undoubtedly continue to change in unpredictable ways. In order for a solution to be scalable, it mustn't be calibrated for any specific set of items.

    We understand this - and were attempting to design our 'maxLoad' algorithm around the changing size of cans and bottles accounting for cooler size. Our aim is to also be able to name the stock quantity needed of each drink type for each cooler location. Although this point isn't really relevant to the base challenge and comes far after a proof of concept that gathers high quality data. Thank you for your prompt response and thank you for redirecting our thoughts back to the exact problem at hand.

    We are putting together a bill of materials right now.
    Is tomorrow evening soon enough to message you the list?
    (we can message you/post a partial list anytime)

    Team Leader for iudicium
    -David

  • Manager   •   about 5 years ago

    maxLoad doesn't seem as helpful to me. you should also know that every cooler has a planogram that dictates what should go on each shelf. Said planogram is subject to change. If your solution can provide better data about what's on the shelf, Coke can take it from there and audit it against the planogram. If your solution were to take it a step further and do the planogram auditing, or begin to recommend changes to planograms in order to maximize transactions, then your solution would exceed expectations.

    Send me your bill of materials. I can't promise I can buy everything but I will let you know very quickly what I can provide. If possible please include links to the products on mouser.com, sparkfun, adarfuit or amazon.

  •   •   about 5 years ago

    @Shane Reiser
    We believe we have multiple decent solutions to identifying drinks removed,replaced,out of stock, attempted guessing at mis-replaced bottles but only one that is cost effective and scales.

    planogram creation for restocking was our goal with statistics based on each coolers individual sales vs history to allow for predictive planning with the ability to alter each coolers planogram per stocking based on each coolers stats.
    We have been calling them 'bays' with text based recommendations(example below)...but a planogram with images is a vastly better description of what we were thinking. Thank you!! We did not know that existed.
    10. Is there a format we can output data so that you can automatically generate a planogram for us?

    We are eagerly watching #coaches for responses to questions 3, 4, 5, 7 from before.
    Working on bill of materials now.
    Team Leader for iudicium
    -David

    Cooler#00000000 on saturday coke classic does not sell more than 50% while sprite, minute maid, and water all sell out. It is recommended that you reduce the # of coke classic bays by *algorithmically determined value* and replace bays 4, 5,6 with: sprite, minute maid, and water.
    Cooler#00000001 never sells more than 50% of any items on current restocking schedule. It is recommended, unless Cooler#00000036 (...nearby cooler list[n]) is being restocked.

  •   •   about 5 years ago

    @Shane Reiser
    I emailed you our BOM
    @all contestants
    Here is our BOM for our best idea http://50.167.142.121/bom3.pdf
    There are some spare parts based on the lot sizes that amazon sells. Does anyone else have a use for the items that have something in the "quantity left for other teams" column?
    [The IP address occasionally changes; I doubt it will within the next 36 hours] If it has changed and you want a copy. Message me your email address or post here and I will give out the new IP.

    "Multiple Submissions: Contestants can submit more than one solution. "
    -http://coolerhack.devpost.com/rules
    We are still working out our 2nd submission's BOM
    We have one with very exact data that costs lots more (modifying cans after production....with RFID stickers....so we are dropping that) and one that is a modification of a weighing system augmented with other sensors. Using a combination of weighing/sensors then combined to produce more accurate data than just a weight sensor alone - as well as robustness to detect if a drink was removed and replaced in an incorrect 'bay'. As well as non drink items being placed in any 'bay'. We are also considering amending our first submission and augmenting with weighing to increase the accuracy of our data; but we don't believe this to be necessary (unless lots of extraneous items are frequently added to the cooler which is something none of us have seen)...we will decide when question #7 from before is answered.
    "7. In Coca Cola's previous failed attempts how accurate was your data? We have discussed perfect systems(100% accurate) that don't scale due to cost or depend on other changes (modifying drink containers is out of the question, we assume) all the way through systems that are economically very reasonable for the data performance. Where is the line between "good enough" and "not good enough" for high quality data?"

    The max price point at production scale:
    Shane Reiser " I have been told in order for the solution to be economically viable the cost per cooler at scale must be sub $20."....hmm. internet looks like the most expensive portion (if cellular) because a final design wouldn't be made out of expensive plexi....but plastic or wire frame....ENC28j60 Ethernet module can be found for <~$3. embedded processor<~$5. Limit switches get cheaper in bulk too, unsure which switch is best yet.
    Does assembly time need to be included? Install time? No matter. Going for KISS.

    This re-raises question 3 and answers portions of it. "3. At each of the cooler locations is internet available/expected(wireless or hardline) or does the solution need to be wireless via GSM(ATT,Verizon) or CDMA(Sprint)?"
    At a $20 price point it will need to be hardline ethernet or wifi. Cellular data costs much more.
    https://www.adafruit.com/products/2471
    http://www.amazon.com/dp/B00A8DDYB6

    We look forward to meeting everyone!

    Team Leader for iudicium
    -David

  •   •   about 5 years ago

    26 gauge solid wire. M/M F/F F/M various lengths.

    Electrical tape

  •   •   about 5 years ago

    Will there will be a projector for showing off data / slideshows? I still have the old school VGA on my laptop... shamefully.

  •   •   about 5 years ago

    @Michael Hug
    I requested 22 gauge and 30 gauge. We should have lots extra. We would be happy to share if either size works for you.
    "Will there will be a projector for showing off data / slideshows? I still have the old school VGA on my laptop... shamefully."

    Good question! No shame! my laptop too has a VGA and then displayport. Almost obsolete and then unpopular.
    We too are wondering about presentation format.

    Team Leader for iudicium
    -David

  • Manager   •   about 5 years ago

    @David - assembly time, install time - not so much. assembly cost, install cost - more important.

  • Manager   •   about 5 years ago

    @Michael Hug - we'll get wire and tape. And yes - there will be a projector that you can connect to wirelessly.

  • Manager   •   about 5 years ago

    @David & Michael - presentation format: more of a science fair format where judges walk around in a group and you demo your solution on one of the coolers around the room. There will also be a giant screen at the front plus two more TVs projecting any slides you want to show and you will be speaking into a microphone during your demo.

  • Manager   •   about 5 years ago

    @All - FYI - I've been consolidating all the questions everyone has asked and the answers I've been getting from the Coke experts here:

    https://hackpad.com/Cooler-Hack-n8XCW3i76nE.
    (scroll down to Q&A)

    Some really good stuff in there.

  •   •   almost 5 years ago

    There are many (hundreds) of different sizes, it so beautiful and thank you
    http://xflori.com/

Comments are closed.