OpenCTD Documentation and Development Workshop (Workshop Update: GOSH Open Science Hardware Event Funding Program 2023)

We are super excited to be among the recipients of GOSH Open Science Hardware Event Funding Program 2023 (Please note: we had a really tight turn around between being awarded this grant and needing to get the workshop organized and implemented, so unfortunately we were a bit slow on getting the forum updates in. The workshop happened June 6 through 8, 2023).

Oceanography for Everyone is an informal collective working on developing the OpenCTD, a low-cost, open-source oceanographic instrument that measures conductivity, temperature, and depth. These measurements allow ocean scientists to assess ocean health and better understand our changing oceans.

The OpenCTD (https://github.com/OceanographyforEveryone/OpenCTD) is designed to be built, calibrated, and deployed by end-users, reducing the barrier to entry for ocean knowledge seekers working in science and conservation. A key component of the OpenCTD design philosophy is that it should be as readily accessible to as many ocean users as possible, without requiring experience in electronics or ocean science. OpenCTD workshops have been conducted with students as young as middle school.

This OpenCTD documentation and development workshop will be an opportunity to bring together contributors to the OpenCTD project with the goals of: updating all the OpenCTD documentation to the current specifications and standards; conducting field trials against known, calibrated commercial CTDs when available; and introducing the local academic community to the potential for low-cost, open-source scientific instrumentation.

This workshop will consist of a two-day documentation-a-thon, where project leads and community members will build and document the OpenCTD assembly process, update the instruction manuals, and solicit feedback on the instructor syllabus for running classroom workshops. Interested students from the University of Maryland Horn Point Lab will be invited to participate in this process.
After going through a calibration process, the newly built CTDs will be deployed in the Chesapeake Bay in conjunction with calibrated, commercial CTDs from the University of Maryland (if available) to provide data validation for the OpenCTD. This will provide an additional opportunity to invite participation for the UMD Horn Point community, including students, faculty, and staff. The goal of this is to not only validate the quality of the OpenCTD construction, but to more broadly introduce the potential of open-source scientific hardware to the local academic community.

CTD’s are the cornerstone of all oceanographic research and an essential component or nearly all marine scientific and conservation studies. One of the many hurdles to large-scale acceptance of tools like the OpenCTD is that formal researchers are often skeptical of the data quality (though, as a side note, among the discoveries made in our 7-year pursuit of an open-source CTD is that many, if not most, of the commercial CTDs currently being used within academia are themselves out of calibration and unreliable). By ensuring that the documentation is up to open-science standards and providing additional validation of the quality of the instruments when compared against expensive and restrictive commercial alternative, this workshop will help build confidence in the potential for open-science hardware in the academic community and especially among oceanographers and marine scientists.

3 Likes

OpenCTD Workshop Update Day 1 - Building the Control Unit, Sensor Pack, and Housing

Day 1 is usually the longest day in the CTD building workshop. In order to calibrate the instrument on day 2, we need to build the control unit and sensor package, test both, and then pot them in high performance epoxy that needs to cure overnight. Were it not for that overnight curing step, we wouldn’t have to frontload the build so much.

Joining us for the OpenCTD workshop is Brian Marx from the University of Maryland Chesapeake Biological Lab, Andrea Schmuttermair from the National Marine Sanctuary Foundation, and Jeff Branson, Chief Technical Mayhem Officer at RobotJazz. I couldn’t have asked for a better team. Also joining use is journalist and photographer Allie Wilkinson, who will be documenting the build process so that we can update the manual with better photographs. Unfortunately, due to a scheduling conflict, my co-founder, Kersey Sturdivant wasn’t able to join us.


The CTD Crew hard at work. Photo by Allie Wilkinson

With help from participants, we identified multiple places within the manual that needed significant updating for clarity and precision. There was a ton of expertise in the room and we also spotted several points where we could shave down the cost and complexity of the CTD build.

After about six hours of focused building, we produced the control unit and sensor pack for two new CTDs, potted them in high performance epoxy, and let them sit overnight to cure.


OpenCTD control unit. Photo by Allie Wilkinson.

3 Likes

Awesome!

2 Likes

OpenCTD Workshop Update Day 2 - Power and Calibration

Day 2 is a much more paid back build day than day 1. With the epoxy cured, we have to install the batteries, finish the housing, and then calibration

The calibration process is new for this workshop. Calibrating CTDs, particularly conductivity, to a level of accuracy acceptable for scientific research is notoriously tricky. For commercial CTDs, users often have to send them back to the manufacturer for calibration, resulting in significant expense while also taking the instrument out of the field. Over the lifetime of the device, that service contract can often exceed the cost of the CTD itself. For knowledge seekers working on a shoestring budget, who may not be located somewhere where mail delivery is efficient and reliable, that can be an insurmountable expense.

In my lab, I would have access to a thermally stable bead bath and precision instruments for comparison, but in the real world, those tools aren’t readily available. Over the last year, I’ve developed a process that used the hot bed of a low-end (<$200) 3D printer, coupled with a Styrofoam cooler and a large ceramic thermal mass (aka, a coffee mug) to reliable hold 25C, allowing us to calibrate the OpenCTD sensors without access to fancy lab equipment.

Documenting this calibration process is one of the major outcomes from this workshop.


It’s goofy, but it works. Photo by Allie Wilkinson.

With the CTDs built and calibrated, there were just a few more steps left to do before deployment day. We took a good chunk of the afternoon to review and revise some of the existing instructions about ballasting and casting, as well as the safety guideline for educators.


The CTD crew with their CTDs. Photo by Allie Wilkinson.

We were also lucky enough to catch one of Maryland’s famous skipjacks, docked up at the Horn Point lab for the day.


Maryland Skipjack. Photo by Allie Wilkinson.

And we got a glimpse a curious juvenile osprey nesting nearby.


Osprey nest. Photo by Allie Wilkinson.

2 Likes

OpenCTD Workshop Update Day 3 - Deployment and Testing

We ran into two major hurdles for deployment day. I had hoped that we would have a commercial CTD available to do side-by-side tests. Unfortunately, no one had a recently calibrated commercial unit available at the lab. That’s a minor problem and I’ve already made arrangements with a local researcher to piggyback on their next deployment to get some comparative data.

Perhaps a larger problem, however, was that air quality on June 8th, 2023 on the US East Coast was apocalyptically bad. Smoke and haze from the Canada wildfires settled over the Chesapeake Bay, dropping visibility precipitously and being generally unpleasant to breath.


Haze over the bay. Photo by Allie Wilkinson.


Haze over the bay. Photo by Allie Wilkinson.

We had a much abridged deployment plan, that was then cut even shorter by a major thunderstorm rolling in. Our planned half-day on the water became a 50-minute round trip out to the closest deep water we could find. A pair of test deployments to 15 meters to ensure that the housings stayed dry, and then a couple of CTD casts to get data to analyze.


Casting the CTD. Photo by Allie Wilkinson


A rare shot of our photographer, Allie Wilkinson, who spent the workshop meticulously documenting the process for the build manual, as well as our Captain, Richie, who got us out and back with all the data we needed. Photo by Andrew Thaler.

2 Likes

When it comes to instrument validation, boring data is beautiful data.

We managed to complete two dry casts (where we send the housings down to the bottom without the electronics to make sure that everything sealed correctly) and two data casts before the weather turned and we had to make a run back in to port. The data came back nice a clean. You can see on the first graph where the temperature probes take a little bit of time to equilibrate, which is why we usually do a 1 minute soak on the surface before descent (and yes, on the Chesapeake Bay in June the water is usually warmer than the air).

Sadly, the nearest NOAA buoy on the Choptank was down, so we could only compare it to buoys in the main stem of the Bay, which were about 2 PSU higher than what we measured (which they should be, the further down stream you go, the saltier it gets). It was a very nice day for instrument validation (and a very gross day to be on the water).

3 Likes

Some deliverables for you!

First off, it’s going to take the better part of the rest of this summer to work through the notes, comments, and expert guidance I received during this workshop. My main goal for this grant was to update the core OpenCTD manual to make it more accessible to more knowledge seekers. The guide it almost 100 pages long and growing. It’s not a fast process.

That being said, it would be a shame to lose the momentum from this last week, so I’ve also been plugging away at a few other tangible outcomes.

Calibration has been the big challenge of the OpenCTD for the entire ten years we’ve been working on it, and we’re finally at a point where I am confident in the calibration protocol. As a complement to the main manual, I just pushed OpenCTD: Calibration and Data Management a standalone guide to calibrating the OpenCTD.

In addition to the guide, I also spent the last day rewriting the woefully inscrutable calibration software we’ve been using from scratch. Anyone who uses the Atlas EZO-EC conductivity probes might benefit from this slightly more user-friendly bit of code for calibrating that system.

4 Likes

Awesome work!

1 Like

So excited to see this.

I remember @andrew.david.thaler talking about this waaaaay back 10+ years ago, and this is bringing back so many memories from my marine biology days. I am blown away by how OpenCTD has grown since then! I particularly appreciate not this detailed thread, but also the extensive documentation that comes with the hardware. :open_book:

Thank you for the update @andrew.david.thaler! Surreal and amazing how GOSH can connect us across space and time. :hearts: :rocket:

P.S. Really appreciate @andrew.david.thaler’s writing, and wanna give a shout out to the Southern Fried Science blog they co-founded!

2 Likes

Oh wow @hpy! Great to hear from you. Thanks for all your support.

2 Likes

Hello Everyone!

I’m very happy to announce that our paper, The OpenCTD: A Low-Cost, Open-Source CTD for Collecting Baseline Oceanographic Data in Coastal Waters, has finally been published! Funding for this workshop was instrumental in getting our paper across the finish line.

7 Likes

Awesome!!!

1 Like

Congrats, @andrew.david.thaler and team! Your work has been a major source of inspiration and has helped me move most of my work into the open source ecosystem. Excited to see what’s next for OpenCTD!

3 Likes

Learned about your project while at Rutgers. Congrats on publishing!

3 Likes