Big Data for Bees Hackathon

This section breaks down the life-cycle of the Big Data for Bees Hack following the steps described in the main guide.

Problem

To protect pollinators, and especially wild bee species, the NL Pollinator Strategy ‘Bed & Breakfast for Bees’ was signed by the Dutch Ministry of Agriculture, nature conservation institutions, agriculture organizations, business and many others. Part of this strategy was the Big Data for Bees hackathon.


Currently initiatives targeting pollination barely have access to relevant information about the local fauna they are supporting, such as food and nesting preferences, or the availability of food plants during the growing season. The aim of the hackathon was to empower these initiatives with better access to basic intel, support them with Big Data skills and give a big boost to the NL Pollinator strategy.

Exploration

Going into the hackathon we explored six challenges through targeted pre events and research:

  1. Decision Support for Beekeepers challenge based on relevant relationships between hyve locations and environmental factors;

  2. Bumblebee observations & flower image challenge: to discover unknown relations within a dataset that contained 18,000 observations of bumblebees, or with other datasets.

  3. Groene Cirkels Bijenlandschap Challenge: an existing initiative generating a lot of data offered access to their data and scientific model, challenging teams to make it more SMART

  4. Green Infrastructure Challenge, focussing on contractors, builders, infrastructural companies and local governments and a decision support tool to support bee-friendly green infrastructure.

  5. Serious Gaming Challenge, to empower the many initiatives that want to do ‘something’ for bees, but lack practical knowledge about soil-plant-bee-relations and bee friendly management.

  6. Farming for Nature challenge: building on top of an existing initiative to engage farmers in monitoring biodiversity, the challenge was to deliver a tool around wild pollinators.

Identification

We sought Big Data and scientific Bee and Biodiversity partners, which we found in Naturalis, the national research institute for biodiversity, Wageningen University and Research and JADS, The Jheronimus Academy of Data Science and the World Wildlife Fund. Representatives participated in the hackathon. 


And then of course we needed some secret sauce: civic hackers that wanted to take up the challenge of making Big Data work better for Bees and pollinators. The hackathon turned out to have a pan European appeal, with participants joining us from as far away as Slovenia and Estonia.

Mobilization

The announcement of the Big Data for Bees hackathon, our starting point for our mobilisation campaign, read that we targeted the following skillsets:


  • Data scientists (Python, R) and machine learning experts

  • Frontend developers (Angular, React, Vue, etc) and UX/UI experts

  • App developers: Android, iOS

  • Data governance experts

  • Subject-matter-experts (beekeepers, ngo’s, pollinator experts, stakeholders)

  • Business developers

  • Open Source enthusiasts

 

Prototyping

You can read more on the results in this recap. Joined winners were the team ‘Happy Bee, happy me’ that worked on a serious game targeting a primary school children and team Hommeles, that showcased some strong data skills for the identification of bee and flower species and ecological relationships. 


The second prize went to an Estonian-Dutch team working on decision support for beekeepers. They managed to build a working prototype, they used additional data such as the Food4Bees indexkaart and the Apiary Map and they shared their code on Github


Other teams built: 

 

Partnerships

Joined winners (serious game/machine learning) received a 20.000 euro prize to develop a prototype. They will be releasing the results later this year.  

The second prize winners were granted 10.000 euro prize money, which was supplemented with  additional financial support from an existing organisation (Food4Bees). They also partnered with VAA ICT, a software company and collaboratively ended up in a European accelerator program on citizen science and earth observation

Part of the team that worked on the farming for nature challenge is still involved and re focused on building a pre competitive and open source infrastructure to facilitate the development of the  monitoring & performance tool. Their first focus - or use case - lies on remote sensing for herb rich grassland (which was dealt with in an earlier hackathon: the rewarding nature hackathon (read more here).

More information

  • Link to Forum (English)

 

©2019 by FarmHackNL. Published under Creative Commons license: Attribution-NonCommercial-ShareAlike 4.0 International  CC BY-NC-SA 4.0