Project goals - please explain

I wonder if anyone can explain the OSBH project goals to me because I feel like I don't get it any more. Few months ago when I found this project on the Internet I was really excited with the idea of solving the CCD problem using the community driven way. I myself am not a beekeeper. Actually I'm a software developer but I'm aware of the problem and kind of apprehensive because of it. I just want to ask you a few questions.

1. What are the actual project goals?
2. What are the project goals in terms of scale?
3. Do you plan to involve professional beekepers?
4. What is your view on the diy'ers participation in terms of data quality? To train an algorithm your data must be reliable but it will be not when collected by different devices you actually know nothing about. It is not a problem when participants are few but when going into large scale there will be significant data pollution obfuscating the picture.
5. How do you plan to analyze the collected data?
6. Is there any kind of a plan or road map for the project? Where can I find it to check if there are any greenfield or support needing areas to participate?

Thanks in advance,
Jakub

Hi @jakub,

Im a little surprised by this to be honest! We have a lot of documentation of our goals (see our website) and we're making good progress towards them.

We don't have the time or manpower to convince anyone the to help out. If you see value, help out. If not, no worries.

You can see a list of to dos for each area by checking the To Do threads for each catagory in our forum.

We appreciate your thoughts, but they are only helpful to the project if you are to helping develop or review something. Your questions can be answered by reading our documetation.

Hi @trizcs

Where is this documentation? I read literally everything on this site but do not actually find answers to my questions. Just read the To Do threads in the Software App section and you'll know what I'm talking about.

Believe me I'm also a bit puzzled. I see a great value in this project. If not I will never spend any time studying this page. To be honest I would really like to participate in it but I just do not get what direction it all goes. I can help you out in the area of software development and data maintenance. I have a lot of thoughts on how it should work to collect valuable data and I know how to implement it.

Below you have my answers to these questions:
1. The primary goal in my opinion is to collect as much valuable bee and environment data as possible and to educate people unaware of this growing problem.
2. The scale is a must. In data analysis only big collections provide the answers.
3. Without professional beekeepers there will be no scale and no big data.
4. This I don't know really.
5. The data must be tagged according to hive inspections. Without this information it will be impossible for the algorithm to learn.
6. I can help out in the area of software.

Jakub

@Jakub - our overview page covers a lot of this. I suppose we don't have all our goals spelled out point by point (perhaps this is necessary) but I feel it's clearly implied in almost everything we do. I imagine this is how you just accurately described our goals.

A little advice from my end would be to just cut to the chase, and if you have proposals, just make them. Our project has a very small team and (frankly) it's a waste of our time to nit pick these things. Collaboration is about building and working together rather than wasting time with pedantics.

We'd be super interested to work with you on software and data if you have proposals. We're only just getting to the point where we have data to work with, so this would be a good time to jump in.

Perhaps a good next step would be to share what skills you have and make a proposal for something specific you'd like to work on?

I'm a Java programmer. 10 years of exp. I work with both backend and fronted. Currently I'm on data mining studies but have no professional experience within this area. That's it.

My idea is to create a really easy to use application that lets people access the data collected in their hives and supplement it with their own input (I mean the hive inspection sheet). This would possibly create a really good data for the algorithm to train. There must be some human feedback of what actually is going on in the hive to make data mining possible in the future. This is only the beginning because the possibilities are actually endless. A mobile app that lets you fill in the sheet online while inspecting apiary in the forest is just another obvious and straightforward idea. It could also collect data from hives that do not have access to the Internet via bluetooth and upload it to the server. It's not rocket science nowadays. But forget the mobile. At the beginning the app should be as simple as it can. It should just have a robust data presentation layer, an inspection sheet form and browsing, filtering, etc. and maybe some visualisation of community data. Because it's first users will probably consist of people with strong awareness of what they participate in and why, no need to egg them on with fancy stuff. In the next step I would go to the mobile app and sensor kits that can collect the data offline for future processing. It would possibly allow much more beekeepers from Europe to participate, because a lot of apiaries are located in the forests where you are completely out of gsm range This is how I see it. Is this somehow close to what you are planning?

This is great info, thanks @jakub.

Please see this thread where @brianzable outlines his ongoing HiveLife project which is doing almost exactly what you're outlining above. @brianzable - please read @jakub's proposal above - do you think your tools could help inform an algorithm used by the OSBH project? I could see this being a very valuable exchange - interested to hear both you're thoughts about how we might make this happen.

Hey @jakub,

Like @trizcs said, I am working on a similar application to what you are describing. The app is written in Ruby on Rails and I am currently redoing the front end in Polymer. I hope to have the app redesigned in the next few weeks to share with the community.

Creating such an app is no small feat. I had an initial prototype of the app running for a little while and ended up shutting it down because the design for this type of application is tricky. Here are some of the problems I ran into;

• Not all beekeepers record data on their hives. Many do, but you need to make something simple enough to convince people to use it. Unless you have a good set of data fields to track, users will quickly stop using the app . The other difficult problem is convincing beekeepers that don't keep records to use the site. I think this can be done if the tools are nice enough and offer enough incentive to start keeping records. This is something I'm not worrying about right now.
• Getting accurate data is difficult as well. The answer to this problem is to automate as much as possible. A good example of this is automatically recording weather data with each inspection. This is easier said than done because many beekeepers will go out to their hives, do some inspections, then record the data later. Free weather API's are easy to use but only provide data within the last 15 minutes or so. If you want historical data, you have to pay.
• Capturing data from sensors is hard as well. You don't know what the data from a sensor is going to look like because many sensors vary in how they work. Another problem is how much of this data you store. Even collecting data hourly for a a decent amount of hives would require a lot of resources and costs to maintain such a site. I'm starting with manual data collection from beekeepers and getting that to a place where most users are happy before I start pulling data from sensors.

Did you have an idea of what sort of functionality your site/app would have to start?

Awesome @brianzable, thanks for the reply.

Could you and @jakub move the conversation into this new thread to keep the things nicely labeled and open to other collaborators? Interested to see your response in there @jakub! @ron might also want to keep an eye on this and send over any thoughts (see above conversation).

Thanks!