WILDLABS Virtual Meetup Recording: Low Cost, Open-Source Solutions

The first event in Season Two of the WILDLABS Virtual Meetup Series was on Low-Cost, Open-Source Solutions. The recording is now available to watch, along with notes and recommended reading based on the discussion. In the session, Alasdair DaviesDavid Lang, and Tanya Berger-Wolf shared their work in 10-minute presentations, which were followed by a lively open discussion and community exchange.

Date published: 2019/03/18

overview

The WILDLABS Virtual Meetup Series is a program of webinars for community members and wider partners to discuss emerging topics in conservation technology and leverage existing community groups for virtual exchange. The aim of the series is to bring leading engineers in the tech sector together with conservation practitioners to share information, identify obstacles, and discuss how to best move forward.

Season 1 of the series took place in late 2018, covering new data collection techniques through Networked Sensors for Security and Human-Wildlife Conflict (HWC) Prevention and Next-Generation Wildlife Tracking, and effective utilization of that information through Big Data in Conservation.

Season 2 will run from April - June of 2019 and will explore different models for collaboration, the low-cost tools these approaches are producing, and what we can do with the data they’re generating. The three topics to be covered are Low-Cost Open-Source Solutions, Tools and Spaces for Collaboration, and Creative Approaches to Data-Driven Storytelling.

There is a lively discussion about possible topics members would like to have space to discuss, so if you have ideas for future meetups please join the thread and share your thoughts.

Meetup 1: Low-Cost, Open-Source Solutions

Date & Time

Tuesday, April 9th, 2019

Main Talks: 3:00-4:00pm GMT / 10:00-11:00am EDT

Additional half hour for discussion: 4:00-4:30pm GMT / 11:00-11:30 EDT

Background & Need

Despite critical advancements in tech solutions being made available to conservationists around the world, many existing tools are cost-prohibitive in the landscapes that need them most. Additionally, those who create low-cost and open-source alternatives to pricey market tech are often operating on tight budgets themselves, meaning they have limited resources for the promotion of their solutions to a wider market. Therefore, there is a need for increased communication around these solutions to highlight their availability, share lessons learned in their creation, and avoid duplication of efforts. Arribada Initiative and OpenCollar are both examples of current efforts in this arena.

Outcomes

The aims of this discussion are as follows: to introduce low-cost, open-source devices for conservation; to describe how they are being used, including what needs they are addressing and how they fit in to the wider conservation tech market; to identify the obstacles in advancing the capacity of these technologies; and to discuss the future of these solutions - particularly their sustainability and how best to collaborate moving forward.

Agenda 

  • Welcome and introductions (5 min)
  • Alasdair Davies,  Founder of Arribada Initiative, Co-Founder of Naturebytes, Conservation Technology Specialist at ZSL, and a Shuttleworth Foundation Fellow (10 min)
  • David Lang, Co-Founder of OpenROV and Open Explorer (10 min)
  • Tanya Berger-Wolf, Co-Founder and Director of Wildbook.org and Professor at the University of Illinois at Chicago (10 min)
  • Q&A discussion with speakers (20 min)
  • Optional ongoing discussion and community exchange (30 min)
  • Takeaways and wrap up (5 min)

Watch the Virtual Meetup

WILDLABS Virtual Meetup Link to Open Source Video Recording

(Click through here to watch the full meetup)

Virtual Meetup Notes

During the first event in Season 2 of the WILDLABS Virtual Meetup Series, we were joined by 88+ attendees from at least 17 countries! Thanks to everyone who participated in the live chat and Q&A - especially to those who were able to stay on for the discussion. Your insightful questions never fail to impress us. For those of you who were unable to join live, we've recorded the session so that you may view it at your convenience. We've also shared presentation notes and further reading suggestions below.

Speaker: Alasdair Davies

Background 

  • 11 years as a conservation technologist based at ZSL 
  • First open source memory: using an old desktop computer as a 10-year-old, I was delighted when I realized I could edit the configuration to make the computer do what I wanted (games), instead of what my dad wanted… 

What does open source mean today? 

  • The freedom to change, edit and customize 
  • Sharing knowledge and celebrating impact together 
  • Breaking down barriers to enable others to succeed 

My path and where we are now

  • Started open source journey working with a device called Mataki - an open source biotelemetry tag developed by Robin Freeman at Institute of Zoology; lessons learned in this work have informed my path since
  • In my three years as a Shuttleworth fellow I have focused on two goals: ​
    •  Develop a sustainable funding model to help support third-party developers of open source tools and solutions
    •  Systematically open source conservation technologies that would offer the most value if made accessible and available to all

Developing open sources tools 

  • Examples of our work: 

  • Your questions fell into these categories (from registration, see this thread): 

    • Sustainability, business models, mobilizing community 

    • Collaboration 

    • Cost vs Quality 

    • Interoperability 

    • Capacity + Skills 

Alasdair open source slide

Arribada Initiative 

  • Open Source Platforms: Arribada Horizon 
    • First completely open biologging platform 
    • Started as a sea turtle tag, now building an App to configure it 
    • Allows others to inherit what you’ve done before them – others in the community now using it to build their own solutions (e.g. Open Collar Initiative)
  • Marketplace: Used by Open Acoustic Devices (AudioMoth); available to all Q4 2019 
    • Group purchasing – use WILDLABS to advertise that an order is taking place; aggregating orders brings their cost down astronomically (e.g. AudioMoths went from over $800 to about $40 USD!!) 
    • Pooled profits – giving 80% of profit back to developers allows for sustainability and further development of tools 
    • Fulfillment and support – making sure logistics are dealt with and support is available; vital for maintaining successful marketplace 
    • Realization that this service provision was the missing piece – NGO needed to act as the glue between the user and the developer 

“It’s only when we get together like this and start to talk about these problems that we can really have an impact as we start to answer and address issues that fundamentally need us to work together as a community to solve.”  

Speaker: David Lang

Background: OpenROV 

  • ​ROVs (remotely operated vehicles) are useful tools for marine monitoring but were very expensive at the time (2011)
  • Wanted one to look for gold in an underwater cave with a friend, so we decided to build one – the vision was to create an affordable, accessible ROV for anyone to use
  • Fundraised through Kickstarter to create original OpenROV kits, which we eventually developed into a more robust product (Trident

Open source 

  • It wasn't feasible to make the Trident product open source in its entirety (although a good portion of it will be open-sourced in the near future)  
  • When does it make sense to open-source?
    • Need to think about what your goals and desired outcomes are - if you become a company, as we did, tensions arise between what you can open-source and what you need to protect as a company to maintain viability 

Platform vs open source - the architecture of participation 

  • Open sourcing is about having open APIs, but being a platform is about whether people are creating more value on top of your solution; tools that are accessible and can be built upon easily are most needed in this space 
    • This distinction is important – while open source is valuable, having a product like the Trident has actually made building and modification more accessible than it was with the open source kit 
  • Many of the challenges facing the natural world are about lack of engagement and a sense of apathy – it’s not enough to get more low-cost tools on the market, we need to think about how to make them more accessible and set up to engage and encourage participation 

Group buying/ advanced market commitments  

  • We've taken a different approach from Arribada’s AudioMoth group buys – instead we pitched the idea to philanthropists, and got them to pre-buy nearly 1,000 units that OpenROV could then give away to conservationists, researchers, students, etc. who were doing interesting things with them  
  • Fill out an application here
  • Lots of potential for these community- based solutions! 

Speaker: Tanya Berger-Wolf

The Problem 

  • We lack critical data on species across the globe 
    • The IUCN Red List currently contains conservation status information on roughly 90,000 species, with a goal to reach 160,000 by 2020 
      • Of what they have, more than 15,000 are data deficient, including some iconic species like orcas, and we’re missing key population information on others including savannah elephants and polar bears 
    • Begs question: can we really scale up monitoring populations on that level, and do it well? Need to be able to answer essential questions like: How fast is the elephant population in Africa declining? How far do whales travel? How many bobcats are left in the world? How many turtle hatchlings survive? 
  • Previous solutions 
    • Great elephant census (funded by Paul Allen Foundation) took 2 years and $8 million USD – that’s not scalable or repeatable 
    • Physical tracking technologies like collars are useful but have critical limitations and risks – even with low-cost tech, you can’t collar every elephant, and implanting chips, etc. can be risky for the tagged animals 
  • So, how to effectively do non-invasive, repeatable, scalable, and cost-effective monitoring of species around the world? 

Our Solution: Images! 

  • We have the data 
    • Images are the most abundant, readily available source of information we have 
    • Images come from many sources, from scientists in the field to camera traps to tourists posting on social media - we just need to design the right computational solutions to collect and extract information out of them
  • Challenges 
    • Overlap of animals in image – how many are there? 
    • Different viewpoints/ poses 
    • Quality, lighting, resolution 
    • Scarring and aging of individuals 
  • What our algorithms can doTanya algorithm slide
    • Find the animals 
    • Tell us where the animals are in those pictures 
    • Classify by species 
    • Identify individual animals (e.g. this is Zippy the Zebra, Willy the Whale!) 
    • Includes information on why the algorithm thinks it found the same animal in two images to aid in human verification 
    • Worked originally for spotted, striped, wrinkled, or notched species, more recently on shape of whale flukes and dolphin dorsal fins, and now on elephant ears 
  • With information on when and where the image was taken, we can start to: 
    • Track individual animals 
    • Population counts 
    • Birth/death dynamics 
    • Species range 
    • Social interactions 
    • Species interactions 
  • About Wildbook.org 
    • Platform for visualizing this data with profiles for individuals 
    • In addition to photos directly uploaded to platform, we use an intelligent agent to data scrape images from social media (e.g. YouTube videos for whale sharks) and use natural language processing to extract info on where and when images/videos are from 
    • Pages also list conservation organizations and research projects that have this individual in their dataset, which facilitates collaboration among projects 
    • Started Wildbook with a couple hundred known individual whale sharks, just surpassed 10,000 known individual whale sharks 
    • Over its 1 year of existence, the YouTube intelligent agent has brought in more data than all of the human contributors combined, 96% of which has been unique data not cited by humans 
    • Now official information for whale sharks on the IUCN Red List comes from Wildbook data 
    • Most comprehensive study on the biology of whale sharks was co-authored by 36 authors that met through the pages of Wildbook 
  • The power of this technology 
    • The Great Grevy’s Rally, 2016: 
      • First deployment of Wildbook was at Lewa Wildlife Conservancy for Grevy’s Zebras 
      • Used Wildbook technology for the first-ever census of an entire species, done by ordinary people driving around Kenya for 2 days (95% of population is in Kenya)
      • Hundreds of people, 45 locations, 40,000 images, 2,352 animals 
    • Round 2 of the rally, 2018 (requested by Kenya Wildlife Service) 
      • 1000+ people, 50 locations, 70,000+ images, 2,700 animals, included reticulated giraffe 
      • These numbers now used by IUCN Red List 
    • Enabled citizens to be a part of creating change in management and policy 
      • Resulted in effective management changes based on population trends across the country from 2016-2018 
      • That data provided the basis for a new species management plan by Kenya Wildlife Service and Kenyan government 
  • Using social media as an information source
    • Bias: spatio-temporal, social media, photo capture, data collection, model selection, etc. 
      • Results in errors in population estimates – need to change statistical models to account for these errors 
    • Risk: social media posts can help poachers track and find targets, have to watch out for this double-edged sword 
      • Working on developing policy around security and privacy for endangered species data (especially geo-tagged, image-based data) 

Take-home message: Machine learning and data science allow us to go from pixels to science, conservation, and public engagement at large scale and high resolution over space, time, and individuals  

Further Reading

Links referenced in the live chat: 

Next Steps

Continue the conversation 

Head over to this thread to add your thoughts to the discussion.

Join our next virtual meetup

Topic: Collaboration Spaces

Date & Time:

  • May 8th, 2019
  • Main talks: 3:00-4:00pm GMT / 10:00-11:00am EST
  • Additional discussion: 4:00-4:30pm GMT /11:00-11:30am EST

Registration:

To join, register here. After registering, you will receive a confirmation email containing information about joining the meeting.