Machine Perspectives on Alien Landscapes: Robotic explorations of Himalayan and trans-Himalayan lakesLatest update September 12, 2019 Started on January 1, 2019
Our aerial and submersible robots
will acquire have acquired imagery, data and a unique perspective on high-altitude lakes in the Himalayas and the trans-Himalayan plateau.
We've successfully completed our expedition, and returned to base where we're processing all our content and analysing our imagery. Our plan was to post day-by-day updates to our OpenExplorer blog, but unfortunately we've returned to discover that this version of the OpenExplorer platform is being shuttered on the 13th of September 2019 (tomorrow!), so rather than post our updates here, we'll be posting them at our organisational blog at www.techforwildlife.com/blog. Updates on our expedition are already available on our Twitter and Instagram accounts (@techforwildlife). We've been informed that in the future, the OpenExplorer blog is going to be used to support National Geographic Explorers, we may cross-post there in the future. Until then, we owe a debt of gratitude to OpenExplorer and the former OpenROV team (now SofarOcean) with a special shout-out to David Lang and Margaret Sinsky for all their support, to National Geographic and the Moore Foundation, to the MPhil CL team at the University of Cambridge and to all the special people who supported this expedition at every stage.
It felt like this day would never come, but it finally has; this is going to be my last 'Preparation' post, because we head out for field work in a couple of hours. We have a team, we have our logistics sorted and most importantly, we have the permits we require to work on the Indian trans-Himalayan plateau!
Our team comprises of four intrepid explorers, namely Nandini Mehrotra, Raghav Srivastava, Gabriella D'Cruz and me (Shashank Srinivasan). We begin our journey by convening in Delhi, and then taking a train and then a car to drive in to Himachal Pradesh. Once we're done with our surveys there, we'll head into the newly-formed Union Territory of Ladakh.
There will be little or no connectivity over the course of our field work, and it's a geopolitically sensitive area, so we won't be live-blogging our expedition. Expect to hear from us again in early September!
As anticipated, we've had a few unanticipated issues. One of our expedition team members won't be able to join us on the tentative dates we had scheduled together due to a change in her circumstances. We're going to miss having her along, and are now looking for another suitable explorer.
In other news, one of Varaha's horizontal thrusters has jammed, and we're trying to have it fixed locally before we throw up our hands and have it replaced. I believe the corrosion-resistant coating on the magnets wore off when we tested it in very silty conditions, and these have subsequently rusted. The other horizontal thruster was also a bit sticky, but some WD-40 and hand-turning quickly cleared that up, while the vertical thruster is completely fine. This has been anticipated and explained in this OpenROV blogpost: https://blog.openrov.com/more-feedback-from-the-field-trident-kickstarter-update-28/,
In addition to losing a team member, the non-functional thruster means that we may have to push our expedition from early July to slightly later in the year. This is No Big Deal, and I look forward to updating this blog as things proceed.
Things are falling into place for our expedition up to the mountains this summer. We've put together a tentative itinerary and have assembled a team as well! We're currently spending most of our time planning out the details, chatting with friends with experience traveling in the region and reaching out through our networks to obtain the necessary approvals to conduct our expedition.
I'm also spending some of my time getting Matsya, our trusty OpenROV 2.8, fully functional. Aside from minor issues that will be resolved with a suitable amount of cyanoacrylate glue, epoxy resin and solder, the major issue is that it lost its starboard propeller on a mission recently. We've purchased another set and the folks over at OpenROV (now Sofar Ocean Technologies) have very sweetly upgraded the shipping so that it gets to us in time (from California, U.S.A. to Goa, India!). Our backup plan if the props don't make it in time is to 3D-print them instead, and field-test Matsya with the new props before our expedition begins. Our friend Kyle Lucas came to volunteer with us for a few months, and used a cloud-based CAD application (OnShape) to design a set of props. Once they're printed, we'll integrate a set of m3 inserts into the cores and then they should be ready for installation on Matsya itself!
Other than Matsya, Arva (the DJI 2 Pro) and Varaha (the OpenROV/Sofar Trident) are in great condition. We've been starting them all up at least once a week to make sure they're functional and that the software is updated.
In a subsequent post, I'll introduce the team and outline our plans!
Introducing our robots
Thanks to the National Geographic SEE Program, we’re very pleased to announce that we’ve just taken possession of an OpenROV Trident! It’s the latest addition to our fleet of robots, and in this blogpost, I’m going to introduce all of them.
First, of course, is our trusty OpenROV 2.8, which we’ve named Matsya. We received this device as part of a grant from the Moore Foundation to explore lakes in Ladakh, as detailed in this OpenExplorer expedition (https://openexplorer.nationalgeographic.com/expedition/tothedepthsoftheladakhlakes)..) While Matsya is a complex device to use, it arrived as a DIY kit which I assembled myself; I’m emotionally attached to all these robots, but *this is the one I have a particularly soft corner for.
Next we have Garuda, our DJI Phantom 3 Advanced. While I purchased this in the US in late 2015, a late delivery meant that our beautiful P3A sat in a friend’s basement in Washington D.C. till June 2017! After 2 years, I was finally able to retrieve and put it to work in India. While the drone is technically obsolete, it’s survived a crash or two, and is still more than capable of fulfilling its primary mission of mapping landscapes. Saying which, the Phantom 3 series are all now just over 4 years old and are not as light or as portable as the newer drones available today.
I was recently asked to conduct a project in North India for which I needed a highly portable drone that could be carried on a plane in carry-on luggage. As the P3A wasn't suitable, I purchased a Mavic 2 Pro, which we have named Arva. The M2P is one of the most advanced consumer drones currently available; it has very powerful onboard AI-based image processing that allows it to autonomously track a moving object, while omni-directional obstacle avoidance makes crashing it very difficult (but not impossible!).
Finally, last but definitely not least, is Varaha, the OpenROV Trident. We received it less than 24 hours ago but have already taken it out for practice dives in both a neighbourhood lake and in the estuarine bay at the mouth of a nearby river. It’s performed flawlessly so far, and is clearly a much-advanced descendant of the OpenROV 2.8. Just comparing the Topside Boxes of Matsya and Garuda is a treat!
All the names of our robots originate from those of characters in Hindu mythology. Matsya (The Fish) and Varaha (The Boar) are both avatars of the god Vishnu that are connected with water and the marine environment. Matsya saved the first man from the Great Deluge, while Varaha brought Mother Earth back to the surface from the depths of the ocean. Garuda and Arva are winged beings; the former is the half-man half-raptor mount of Vishnu, while Arva is one of the ten horses of the moon god Chandra.
I spent part of 2008 living by the shores of Tso Kar, a high-altitude lake in Ladakh, India, studying the behaviour of the black-necked crane (Grus nigricollis). These beautiful birds inhabit the upper reaches of the Himalayas and trans-Himalayan plateau, spending the summer months nesting in the wetlands surrounding high-altitude lakes. During the months I spent studying these birds, I observed this lake from a number of vantage points; however, due to the altitude, there are a limited number of perspectives a human can obtain on this habitat.
Today, using modern robotic technology, it is possible to study a landscape from a number of machine perspectives that a human would not be able to attain. For example, we can use underwater remotely-operated vehicles (ROVs) to explore the depths of water bodies that human divers would find too risky, and similarly unmanned aerial vehicles (UAVs) can be used to obtain overhead views where a helicopter cannot be deployed.
For the past two years, I've been running Technology for Wildlife, an organisation devoted to amplifying conservation impact through the appropriate use of modern technology. We've been using an ROV and a UAV to explore water bodies and terrestrial ecosystems in India for a number of projects. We've already tested our current ROV up in the Himalayas.
This summer, thanks to a National Geographic Early Career Grant, we're going to be heading north, to explore a set of lakes in the Himalayas and trans-Himalayan plateau using robots. While the primary objective is pure exploration, to gain a unique perspective on these alien landscapes, our secondary objective is to look for signs of plastic contamination of these otherwise pristine environments. If, as we suspect, we do find such evidence, this will help determine conservation actions to protect these beautiful lakes and the ecosystems surrounding them.
Contribute to this expedition
Thank You for Your Contribution!