Supply chain and technology

sam@earle.com's picture

This is my first time on here since PASA, so am trying to figure it out. Here is what I am interested in:

I beleive global food system efficiencies such as technology can be used to address problems of local food economies and make them a larger part of the local food economy as a whole.

I beleive this will be a non linear system such as the internet (and this networking tool), as opposed to the old supply chain linear model. "Distributed supply chain mangment"

I'm trying to build that software and system that local farmers can use to connect, first to institutions and schools, and eventually to large and individual purchasers (wholesale and consumers).

If anyone would like to discuss, please get in touch with me. I've been working on this for more than a year and have made some great progress.

Cheers

Groups audience: 

Group content visibility: 

Use group defaults

Attachments: 

 

Comments

 Hi Sam!  I have to tell you,

webadmin's picture

 Hi Sam!  I have to tell you, I am pretty interested in the same problem. Right now, I am working out some code using open source Erlang programming language to do what you are talking about. 

By designing for interoperability first, we hope to make it easier to connect with others systems/data sets/approaches. Software is "situational" so we've worked to make a system that can be adapted to unique local needs. We'll be sharing our full code base under an open license soon. Would love to know more about your approach. Thanks

Hi Sam - how soon will yours

sam@earle.com's picture

Hi Sam - how soon will yours be release?

I'm coming at things from a more traditional end, have partnered with some softwre providors who work with the likes of Aramak, etc... with the hope to be able to retool it for a non linear model.

But, I'm waiting for a large institution to say, "if you build it, we will come"

Logistics group may be interested as well

Casey Hoy's picture

Hi Sam,

I'd suggest posting this one to the logistics and perhaps the modeling and data mining groups as well.  David Reed and Trevor Clatterbuck are two who have posted to those groups on similar interests.  I've also recently made some good contacts on this in the Fisher College of Business at OSU, with John Saldanha, who works on logistics and routing but is interested in the larger issues you raise.  Look forward to hearing more on your work!

I can't remember what the

athensown's picture

I can't remember what the distribution entity in the report is called, White Dog, something Dog...can you picture this open source template that you describe as serving their needs and the current services of a resilient local food system that they bring to the table?

White Dog Community

sam@earle.com's picture

White Dog Community Enterprises. They also had a sub-operation called Fair Food. They have since split. Fair food is farm to consumer direct, and consulting to restaurants (http://www.fairfoodphilly.org/) and the other is just a restaurant and social entreprenuership incubation (http://www.whitedogcafefoundation.org/).

I can picture it VERY well, and have the system all designed..... in my head. Yes.

Market Maker

athensown's picture

I can't really shed much light on this.  I am listed on it...that's about it.

DVRPC report

athensown's picture

<p>I can't remember what the distribution entity in the report is called, White Dog, something Dog...can you picture this open source template that you describe as serving their needs and the current services, of a resilient local food system, that they bring to the table? </p>

Hi Constantine: Yes, I can

webadmin's picture

Hi Constantine: Yes, I can imagine a way to use the engines we're creating to address resilient local food systems. Feel free to contact me at samuel.rose@gmail.com if you want to talk about it (or continue here if you'd like)

Casey, I was talking recently with Trevor Clatterbuck about creating some ways for Fresh Fork system to talk more successfully with other systems. This is the stage that Trevor sees he is at. Given that many of these systems are going to emerge, (and rightly so as there will usually be a software developer or team of developers working directly with a community that has specific needs) getting them to talk to each other in standard ways will be extremely important in my opinion. So much so that I'd recommend people making these systems start out making an API (or way to allow data and programs to interoperate) *first* before making any other part of the application. It's my prediction that those people who do create systems for interoperability will survive and thrive. Those that make closed sysems will likely reach a ceiling and struggle to maintain after a certain threshold is reached (the "Red Queen" effect from Lewis Carrol's  "Through the Looking Glass"  "It takes all the running you can do just to stay in one place"  http://en.wikipedia.org/wiki/Red_Queen )

DVRPC

athensown's picture

As a chef I do not want deliveries from ten different farmers.  That's ten times the operational overhead of a single delivery. Phone calls to order, someone to stop, receive, and put away the order, invoices, and checks written and mailed and accounting to maintain.

Do you see the place for community level distribution in a resilient local food system?

I mistakingly thought White Dog or Fair Food was visioning distribution like this.

Hello Everyone, I'm sorry

joshuak's picture

Hello Everyone,

I'm sorry I've jumped in late on this discussion, but I have something to contribute.

I've been working off and on with the xTuple ERP system to make it farm-friendly.  Here's a link to xTuple:

www.xtuple.org

xTuple is a FREE ERP/CRM system.  It was originally designed for manufacturing, but the company took a subset of their enterprise / commercial app and released it as open source.  This app has everything you'd want in an ERP and Accounting package: GL, AP, AR, Inventory, Reporting...

It would be relatively easy to create data feeds into and out of this product to enable EDI and other distribution mechanisms.

I will shortly post the modifications I've made... at this point it isn't much, I've just imported the USDA nutrient database into the inventory system.  I'd like to share this in such a way that we can cooperatively build a complete chart of accounts for the typical farm producer and consumer.  If someone would like to help, I'd like to follow a traditional open source model where the product itself is free and third parties are free to sell services around it.

Thoughts?

Cheers,

JK

Have you guys taken a look at AMQP?

joshuak's picture

Ok, I know this isn't the place for low-level technical discussion, but have you guys checked out AMQP?  AMQP is a message passing infrastructure that in this case can help achieve some of the things you guys are looking at.  For example, a web site could very quickly give you the current inventory levels of tomatos, for example, for a given area - if all the farmers in that area were connected and did not mind sharing this information.

I know how to design these types of networks - I'm currently doing a similar project for a similar but different type of network.  For more information:

1) Article I wrote for Linux Journal last year:  www.octoberridge.net/http://www.octoberridge.net/AMQP-Article-Final-WithFigures.pdf

2) Apache AMQ (QPid) site:  http://qpid.apache.org

Cheers,

-JK

Price reporting as well as inventory

Casey Hoy's picture

Joshua, some of the members of our project are working on a price reporting mechanism through the network, and your suggestion for an inventory reporting system could dovetail nicely with that.  Sam, i think Stan was going to talk to you about this soon.  Perhaps there's a way to pull Joshua's and Stan's ideas together here.  Thanks for the contributions Joshua, great ideas!  Casey

Excellent, thanks Casey.  Is

joshuak's picture

Excellent, thanks Casey.  Is there a technical mailing list or forum where more technical discussions can be had?

Tech group

Casey Hoy's picture

Joshua, The logistics and GIS, Modeling, Data Mining groups are probably the closest groups to what you're describing at this point.  Most of the groups have a fair number of interested readers but don't necessarily function as an active working forum at this stage.  Your communications have sparked a lot of interest and maybe even some good working connections so I'd suggest you keep it up.  However, the focused working forum is what we hope emerges over time and it's pretty easy to launch a new group if you'd prefer with a title that would make it clear it's for technical discussion.  I bet Sam would join if you create one on open source programming, and perhaps others.  I'm not as far up the curve as you guys but can see where connections need to be made sometimes. 

One additional comment for the thread:  Local Roots market and cafe in Wooster, Ohio, developed their online ordering system using open source code, from a co-op in Oklahoma I think.  They can probably tell you more but it might be another starting point.

Constantine, what if we could

joshuak's picture

Constantine, what if we could reduce that operational overhead?  Would you then be willing to take deliveries from ten different farmers?

JK, - the chef takes a

athensown's picture

JK,

- the chef takes a comprehensive approach to building rural resilience. 

- she wants to know, considering peak oil and climate change, the CO2 footprint of separate versus combined deliveries.  Is there a place for a community delivery network, in addition to farmers markets, etc., that takes the profit from networking the goods, and invest it in the communitry.  It can be a nonprofit cooperative if that what folks want, and it can also be a highly dynamic community owned for-profit. 

- this chef knows that farmers are not getting paid for the part that constitutes the delivery, and are mostly  pricing for palatability and minimizing shell shock, and not what the product should be costing, especially when you consider eating is more important than legal advise.  i'm telling you, this chef is on the ball.  And she wants to point her dollars towards solutions.  Comprehensive solutions. 

So, from a community perspective, what combination of mechanisms make the most sense for everybody?  Cuz, only the one that does, makes sense to pursue.  The best system is what she wants to finance with her food dollars, not just a myopic part of it. 

- As she sees it, the delivery has a cost, whether the farmer or a hired driver does it.  It difinitely has a CO2 cost.she also knows the food cost leaving the farm should be higher, and even higher as fuel goes up. 

-I told her I do not have all the answers, but we will try to find them.

Operational overhead It would

athensown's picture

Operational overhead

It would seem, as a community, that we try to minimize operational overhead in every way that makes sense, and try to strengthen systems that do the same.  It seems that what is being proposed here is a step in that direction. 

Some thoughts

- A product availability board is just as useful to a distributor as end user.  

- It will always be up to the farmer how to price for different pick-up or delivery arrangements. 

- A distributor can be quite helpful in posting buying needs in the crop planning stages over the winter. 

- Any way that this system can total seasonal purchases, as well as track per week product flow, at different times of the year, will be helpful to restaurant, farmer, or grower.

- The distributor can be a very powerful ally at getting the farmer increased profitability.  For example, increased culinary knowledge on its staff can cooperatively facilitate needs of both chef and farmer.  Marketing skills on the distribution team can highlight the grower, likely increasing the grower's sales, as well as all food service operations to which it delivers.

- it may be worth considering that a distributor is strategically poised to catalyze or generate a local currency.  There is a portion of its closed loop of buying and selling that does not necessitate bank participation and can grow naturally amidst pre-critical mass realities.

Sam, I have developed  a spread sheet pricing template as part of tracking cattle sales, not as in how many head, but every cut; This helps track profitability, and is a part of the overall tracking and documentation, yet unrequired by law, of every piece of beef that we handle, pasture to plate, without an implant. Do you envision that such a template, could be user added for anyone to use.  I guess it can be added as a file to a shared resource library for anyone to download and or improve, and use as they choose.

Second, I am not so optimistic about the parts I am doing on Quickbooks.  I am at least trying to create any list in excel, so those should be broadly usable   A common product numbering framework highly interest and teases me. anyone working on that?

 it will be helpful if an envisioned template walks me through a second currency, mainly financially accounting for dual currencies. 

 Altnernative currency

webadmin's picture

 Altnernative currency accounting would be tied to and associated with the resources that the alt currency represents. So, you'd need a resource management system that can track different resources based on the nature of the resource.

 Constantine, you talked

webadmin's picture

 Constantine, you talked about a "common product numbering framework" can you talk more about what you mean there? Like a distributed system to that can place unique ID's on products that many others can use?

Bar code generator

Casey Hoy's picture

How about freeware simple barcode label generating software that can be used by anyone to create a digital trail from the marketplace back to the source.  There's a specialty coffee company that is doing this, adding additional labels at each step in the journey for their beans from individual orchards to the bag, even for blends.  Doesn't seem like it should be that difficult and would be useful in marketing farms as well as food.

re barcodes

sam@earle.com's picture

I'm looking into this a lot recently - there are some very simple divices and programs out there. Specifcally, iphone and ipad applications. Put an ipad in every farmers hands, somple barcode reader and network connectivity out of the box. Also some more technical divices, but have specifically seen, local small farmers don;t want too much technology.

Josh, thanks for posting the

webadmin's picture

Josh, thanks for posting the link to xtuple ERP system and AMQP

 

I think there's an emerging need for at least 2 different ERP style systems:

 

One ERP system addresses tradtional mass markets where volumes of production are larger, and crops produced are limited to a very few at the most per-producer. This system mostly connects producers to distributors to consumers. The software is open source, so this basically opens up traditional infrastructure to more people. However, the *market* is the same as it has been.

A second type of ERP system addresses many scales, many niches, in an adaptive and flexible way. it becomes more useful for many smaller producers who work together. And, it facilitates transactions and exchanges that nothing else out there really helps with right now. This second type of system also *reports* on these new types of exchanges. 
 

The system we're working on now does both the first type and second type. The participants choose. 

I like AMQP, although I would like to offer a *plurality* of ways of passing messages in systems. So, RESTful HTTP, SOAP, XML-RPC, sockets, and AMQP. The reason why I want to support multiple standards is that there are already a lot of people out there, active in managing data about local food systems that are already employing multiple standards. As long as someone is employing any standard consistently, they should be able to connect with others and inter-operate.

RE: Common Product Numbering Framework

joshuak's picture

Sam, you noted an interesting interpretation of Constantine's common product numbering framework.  I thought about it a different way... basically, if we have 5 local farmers all using QuickBooks, if they grow tomatoes and oranges, how do we aggregate stats from all of the farmers?

With xTuple I started with the USDA Nutrient Database.  This database has comprehensive data on tens of thousands of food items, and each item has an ID number.  I used this ID number as the "Item Number" in the inventory database.

www.ars.usda.gov/main/site_main.htm

Altnernative currency Numbering

athensown's picture

Nature of the resource

- does that include the food production as well as a service such as distribution

 

Numbering that resource

- beef, and meat in general, has a widely used national numbering system, from the National Assoniation of Meat Packers.  I am trying to incorporate it into the creation of item numbers, and exploring using QuickBooks item 'class' to further track ranch of origination, kill date, and rail weight.  The combination of those three items forms the ID of the file of that animal.  Interestingly, this numbering system list a separate set of UPC numbers recommendations.  Go figure.

Are we able to attach files here?  Is there a spell check?

 

 

 Constantine: there's no

webadmin's picture

 Constantine: there's no spell check, sorry. Although, there could be a way to implement adding attachments to comments (right now you can only attach to posts). For now, let me suggest that you create a new post for your attachment, and post it to the same groups this post is sent to: 

 

 

 

Quickbooks is nice software. I do agree that they do not likely give you much flexibility with adding numbers, ID's etc

I have been using Apache ofbiz http://ofbiz.apache.org/ for my own business accounting, and I think this could be a great base from which to extend and add in required fields and ID's etc, as it is open source software and more flexible in ability to adjust than quickbooks and other commercial products.  Some might also find http://www.opentaps.org/ useful as well. 

I'd recommend that where folks are looking for software, that they do a needs assessment (mapping out what they need as they see it) and a constraints assessment (mapping out the constraints, such as your time, your budget, etc)This will really help guide you towards what will work for you, given that there are many choices, and many unique local conditions for people.

 

Similar Project Here in Southwestern Ohio

RStewart's picture

We have something similar going on here in Southwestern Ohio.  In Cincinnati we have Findlay Market which is designing it more as an auction house and then a group of individulas working with The Ohio Valley Food Shed ( http://www.ohiovalleyfoodshed.org/) group to design both a front page and communication/forum system that would allow producers to inform the buyers and the public about their business (consider this similar to something like Market Maker or Local Harvest) with an interactive buyer/seller interface that would allow farmers/producers to sell and chefs and comissaries to purchase food as well as reserve crops.

 

Its in its iinfacy right now and a good chunk of it is still in the discussion stage.  We (open community as opposed to the initial Ohio Valley Food Shed group) just had out first meeting to begfin discussing it this past Monday.

 

same in WI, NC, etc.

Heather Hilleren's picture

 Similar software project in Wisconsin funded by the National Science Foundation.  A group in North Carolina is using it for exactly what you are talking about.  Better yet, it's free. (www.localdirt.com

It works for wholesale and individuals, just like you were thinking.

Working on putting barcode functionality into it.  Casey, do you know what the freeware you mentioned is called?  I'll check into it.

Bar code idea - still an idea

Casey Hoy's picture

Heather, I don't know of any freeware that does what I was suggesting, but I think it would be used a lot if it is developed well and in an open source environment. 

Heather, that's an ineresting

joshuak's picture

Heather, that's an ineresting site!  How is the site linked to the farmers - do they have to enter everything into the site, or is there some feed that comes from their accounting package?

 The site is bottom up

Heather Hilleren's picture

 The site is bottom up instead of top down.  It takes online orders, creates a purchase order, automatically updates the farmer's price sheet, and now we're developing the API to connect up with their accounting systems so they don't have to do data entry.

For better or worse, we are very big on privacy.  All information on the farm can only be entered and updated by that farm.  They can also choose their privacy settings to adjust how much info they want on the internet.

Good idea? Bad idea?

Heather, does your group have

joshuak's picture

Heather, does your group have a website?  I'm doing something similar (with regards to the API -> Accounting system) and it'd be great to compare notes and ideas.

AMQP / Protocols

joshuak's picture

Heather, this might interest you in the way of API's that talk to accounting systems.  Just got your email, so we can discuss further offline (or in a more appropriate forum) if you'd like.

Sam,

I agree 100% on being interoperable with as many different protocols as possible.  In this regard, Let me share what I've been doing on web projects that aren't related to Farming.

The bulk of the repetitive, tedious code is modeling various objects in all of the end-point systems.  For example, an accounting system might have 60 "objects" such as Journal Entry, Invoice Header, Invoice Item, Inventory Item, etc, and these objects all exist as tables in the database.  If you are constructing applications in Python and Java that need to talk to this system, you have to duplicate those objects across the board.

I am constructing a system, based on Django (the web framework), that will take a database schema and give you workable, serializable objects in Java, Python, and C++.  The "object generation" utility uses Django to inspect the database, then it takes the resulting Django-Python representation and creates a specification file for Google Protocol Buffers.  It then uses the Protocol Buffers compiler to generate native objects in Java, Python, and C++.  These objects are serializable and ready for transport.

That part is working - I'm continuing by extending this to generate objects that work with a particular (open source) C++ database library.  My original goal with this exercise was to be able to connect Django-based web sites to the xTuple ERP system, and to be able to secure xTuple with SELinux (hence the C++ parts).  But that's not Farm related.  :)

How this ties in with what you said: AMQP is more of a transport-layer protocol like sockets.  You can take any higher-level protocol (XML, REST, JSON, HTML) and send it over AMQP just as you'd send it over a bare socket. 

The advantage with AMQP is that it's very flexible in specifying who gets the message and how replies are made.  For example, a website could send a request to an AMQP broker: "I want to know the quantity of tomatoes in Knox County, Ohio"  For all farmers in Knox County who are a subscriber to this broker, their accounting system would be listening for such requests and be able to report an immediate "available" inventory, inventory amounts reserved to fulfill contracts, etc.  The web site would not need to know the IP address of every farmer, or need to keep a list of "connected" farmers, etc - it would just ask the broker and the broker would take care of all of that.

Of course, the Protcol Buffers objects are not the same as JSON, etc. that you noted above, so under a system like I outlined in the preceeding paragraph we'd need to come up with a bridge, or create another connector based on the database schema.

Best,

-JK

Barcodes and Smart Phones

RStewart's picture

The Google and iPhone both are capable of reading barcodes.

If the iPad has photocapability than even the iPad would be able to do it.

 

Our group has been in contact witht he folks that developed LocalDirt and they are building an ap to work with LocalDirt.

 I talked a bit with

webadmin's picture

 I talked a bit with Constantine today about the problem as he sees it.

 

Basically, I communicated my opinion to Constantine: you would not need iphone or any phone to do barcode scanning. All you need is a computer and low cost USB reader that actually enters code in as keyboard would type it. Works with any PC operating system. Arduino could be used to make a wifi-enabled mobile scanning device (using same low cost scanners). RFID is another option. This is all do-able now.

Numbering system would be  easy, even in an environment with multiple numbering systems.  Choose a replicable standard for numbering, and this should then be able to talk to others systems (you could use UPC, or other open standard.)

 

JK:  We should talk sometime. We have some ways of making what it looks like you are working towards easier to do without changing what you are doing.

Leafy greens, Auto-ID, traceability, and supply chain management

choy's picture

I worked with Ohio leafy greens growers last year to address the Produce Traceability Initiative (PTI) promoted by the Produce Marketing Association and United Fresh Produce Association with the goal to provide a common framework for produce growers across the supply chain (field to fork). As has already been discussed in this thread product item numbering needed to be standardized.  PTI adopted an embedded 14 digit Global Trade Item Number (GTIN) with a time table for adoption by growers, shippers and receivers covering several milestones to be completed by 2012. The initial piece that we worked with was tracking produce from the field to the packing shed utilizing WI-Fi, GPS, RFID/bar code technology. Our goal was to make the solution open-systems and adaptable to a variety of back-end accounting systems. We have a great resource here in Ohio for this technology through OSU Electro-Science Laboratory (OSU/ESL). The project was submitted to the Third Frontier Sensors Program for funding, but is not on the short list for approval. Let me know if this helps.