This is an old revision of the document!
So you want to build a CrowBox and try your hand at training wild corvids? You're in the right place. Here are the steps you'll need to follow to get up and running:
Let's go over those in a little more detail:
You have basically three choices:
Ponoko is the easiest as they provide easy-to-use guidelines and ship everything to you, but note that the biggest price breaks come if more than one person orders at once. We’re not making any money from any of this, so if you can, combine your Ponoko order with a friend's and you'll both receive free shipping! (We don't make any money when you buy CrowBox parts from Ponoko.)
If you want to cut the materials yourself, chances are you already know about the necessary file formats, etc., but do take care to match the plexi thickness (3mm) and avoid bright colors. You can download all the necessary files as Open Source flat vector geometry files for laser cutters from the Bill of Materials (BOM).
We don’t have any validated evidence that the birds don’t like them, but generally corvids are scared of new things so matte browns and greens seem sensible. If you’ve got questions, hit up the forum.
Your last option is to find someone who has a laser cutter. MakerSpaces and HackerSpaces frequently do, are usually not too far from you, and are generally filled with awesome, community-minded folk who can help you out. If they don’t have classes, they’re nevertheless great places to start in figuring out what it takes to get the designs turned into a real-world housing.
While you’re getting the housing in order place your order for the electronics. We’ve got a full Bill of Materials (BOM) listing everything you need. You can source directly from EBay (often the cheapest source) or Amazon (especially if you’ve got Amazon Prime) but you can also often find components at places like Home Depot or your local hardware store.
Note that many vendors offer some of the smaller parts in bulk only, requiring you to buy more than you’ll need to build a CrowBox. If you’re habitually making cool hardware solutions this should suit you fine, but if you’re not, see the aforementioned hacker / maker spaces. They may be willing to sell you parts piecemeal.
A special note on cameras: right now the CrowBox software contains code stubs to control cameras, but integrated cameras are not supported as part of our initial release. If you want to record video / photos of corvids using your box (and we totally encourage you to do so!) we’ve tested the following as standalone solutions:
More on future features we’d like to include here.
This can be tricky and requires a lot of judgement on your part. The good news is that you can change your location based on the behavior of the corvids over time. The bad news is you may have to. ;)
Lucky for you we’ve prepared a complete guide on preparing your site for the CrowBox. The information in this guide can dramatically speed up the agonizing process of waiting for Corvids to find and use your new CrowBox.
We’ve got a complete hardware assembly guide video series. If there’s demand from the community for a written guide we’re totally open to someone transcribing the videos and converting them to this purpose! We’ll do our best to support people as they build their CrowBoxes, through the forum.
You should also download and print the paper tools to go along with the video assembly.
Once you’ve got your hardware put together, configure the software and set it up for your first phase of training. More on that in our software configuration guide.
This is likely to be the hardest part of the process, because at first what’s likely to happen is… nothing. It’ll take a while for the local corvids to work up the courage to visit the box and start interacting with it. Be patient, observe, and take notes. If you have a camera, review any footage to get ideas on what changes you might want to make, and wait.
If after a week they’re still not approaching it, try changing one thing and see if that helps. Wait another week, watching closely to see if the change has had an impact. Repeat the process until you see birds interacting closely with the box. This is where the forum is your friend; waiting is not easy, and it helps to get suggestions and ideas from other CrowBox maintainers.
After you’ve seen the birds interacting with the CrowBox for a while you’ll start to get a sense for their confidence and comfort in interacting with it. When you think the time is right, move to the next phase and then back off to observe again. From what we’ve seen there’s huge variability in how long a particular corvid population takes to adapt to each phase, and often you’ll have to roll back a phase or two due to external events.
For example, at once point a hawk moved into our training area and all the corvids totally ceased to interact with the local CrowBox until the hawk moved on a week later. We didn’t realize what the problem was until we saw it had gone.
In another case, the local crow population we were working with totally refused to move into the final stages because the blue jays were happily grabbing all the nuts and caching them… in the tree where the crows liked to sit. These crows were so clever about robbing the blue jays that they never had to deal directly with the CrowBox.
The moral here is that there is no hard-and-fast rule. These are smart birds, but they’re also pretty unique to each other and to their environment. The goal of this project is to figure out what works and what doesn’t in training the broadest subsection of them. To that end, the most important thing you can do is…
We’ve built what we hope is a rugged, relatively approachable design to interacting with corvids using nuts and coins. But the hard work is really yet to be done - working with the corvids and the box to figure out how to get them to use it in it’s final format: fetching their own coins for peanuts.
To that end, the more carefully you document your results (including # of cycles per phase, length of time in each phase, number of corvids using the box in your location, the weather, notes on the environment, etc. etc.!) and share those with the community, the better! The forum is an ideal place to post these as well as to ask questions, compare notes, and suggest changes to other people’s placement and protocols.
(Note that another feature not yet included is a cloud-based data logger to automate data collection and facilitate sharing. More on future features we’d like to include here.)
The future of the CrowBox is now in your hands. It’s a lot of work, filling the machine with nuts, taking notes, observing and altering the protocols based on what you see, and generally just nail-biting as you wait for that exultant moment when the first bird lands. We’ve found the work more than worthwhile as we’ve developed the CrowBox, and we are thrilled you’re joining us on the most exciting part of this process: actually testing it. Together we can use the CrowBox to learn more about these amazing species and, possibly, help change how human beings approach interacting with them.
See you on the forum!