Enjoy "remote snorkeling" over extremely long distances with real-time panoramic video.
Equipped with a solar panel for perpetual use without charging.
Automatically navigates on programmed routes using GPS.
Our ultimate goal in developing the SurfaceWalker is to create a platform to visualize change of natural environment. We are currently working on a collaborative research project with a university to develop OpenCyanoMap, which creates visualizations of the concentrations of cyanobacteria in lakes. Our end goal is to create the conditions that will allow people to gain an accurate understanding of the issue of the decline of coral reefs. We intend to accomplish this by improving the SurfaceWalker's operability at sea, and by developing OpenCoralMap, a feature that will create visualizations of the extent of coral bleaching, as understood with the use of images filmed by the SurfaceWalker.
The Picam360-SurfaceWalker is an unmanned surface vessel (also known as an aquatic drone) that can be operated over extremely long distances, allowing you to visit far-off places, even from the comfort of your own home. A panoramic camera is mounted on the bottom of the drone. You can watch the video being streamed from this camera in real-time on the screen of your smartphone, and freely change the viewing angle, all while operating the easy-to-use drone. This means that you can experience the fun of what we call remote snorkeling at your convenience, even in places that are difficult to access, and even if you're not a strong swimmer.
In Access Point mode, the SurfaceWalker will allow you to connect to it directly with a smartphone. What's more, by using it with a USB 3G dongle (which is for connecting to the internet while outdoors), you can operate the SurfaceWalker from any place where there's an internet connection. As one example of the possibilities, this will allow you to have fun viewing the beautiful undersea scenery of a far-away place, from the comfort of your own home.
The SurfaceWalker can be operated from a smartphone, tablet, or PC, all while you view the images being streamed from its camera. It can also be controlled with the included transmitter. (When using the transmitter, the SurfaceWalker can be operated within the range of the transmitter's signal. The signal range is approximately 500 meters.)
Real-Time Panoramic Video
With the real-time, 360-degree video from the panoramic camera mounted on the bottom of the drone, you can watch the action from any angle. The SurfaceWalker's proprietary image compression technology enables streaming at a bandwidth of 300Kbps.
This video can be viewed on a smartphone, tablet, or PC. As the video is 360-degree video, you can always look in the direction of your choosing, even if your network is experiencing lag. Normally, video filmed from an aquatic drone can be unstable, due to the hull swaying on the surface of the water. However, the SurfaceWalker's camera is equipped with an attitude sensor, which corrects for this motion, making sure that your view remains steady. This capability is specific to the SurfaceWalker's panoramic camera, and is something that ordinary cameras don't offer.
Solar Panel and Battery
Equipped with a 100W solar panel, it can continue to operate on the power generated during a clear day alone.
Will it still work in poor weather conditions? You might be worried that the ship won't be able to come home safely. To prevent this, an additional battery can be added to SufaceWalker. Rest assured that it will keep moving even in poor weather.
Unmanned Autonomous Navigation (To Be Developed)
The SurfaceWalker is equipped with an autonomous navigation feature that allows it to navigate along an assigned route without additional human input, using GPS. When equipped with a camera and measurement devices, it can automatically collect data about what it observes both on the surface of the water and under the water. At present, the autonomous navigation feature uses a simple control screen, from which users directly input the desired coordinates. However, we plan to develop a UI that will display the SurfaceWalker's current situation, and will allow users to control the SurfaceWalker in a more intuitive way, by pointing to a specific location on a map.
Observation Data Mapping Feature (To Be Developed)
We are developing features called OpenCoralMap and OpenCyanoMap, which automatically upload underwater images filmed during autonomous navigation and data collected with the drone's measuring devices to a server, and display the collected data on a map. Our goal in developing these features is to help find solutions to environmental problems. The development of these features is the end goal of this project. For more details, please see the section of this page entitled The Final Goal of the Project.
Picam360 : The SurfaceWalker's Proprietary On-Board Panoramic Camera
With regard to the changes in hardware in recent years, there is now more hardware than ever that will allow even individual users to become developers. Coinciding with the falling cost of connecting to the internet, we've seen the popularization of mobile devices with large displays, such as smartphones, as well as the appearance of single-board computers like the Raspberry Pi, which provide the functionality that developers need at a low price. Meanwhile, with the growing disparity between the relative pixel densities of image sensors and displays, the resolution of image sensors is now well beyond what is required for most of today's applications. As a result, there's a growing movement to make better use of image sensor resolution, which is worth watching. Notable examples include light field cameras, which allow users to freely alter the focus of a photograph even after taking it, and hemispheric panoramic cameras, which have viewing angles greater than 180 degrees. There are even signs that a trend toward fully spherical panoramic cameras, which are made up of two hemispheric cameras, is set to take hold. In the midst of these changing conditions, we have begun the development of an open-source panoramic camera called Picam360, in the hope that the existence of an open-source system with a focus on "Smartphone x Network x Raspberry Pi x Panoramic Camera" would enable new value to arise from the ideas and insights of Makers from around the world. Networked cameras, which combine "Network x Camera" technology, are already in existence. These cameras have needed rotators in order to compensate for blind spots caused by the cameras' limited viewing angles. When added to the constraints imposed by data transfer latency, this has often resulted in a discrepancy between what the user of the camera is seeing, and the image produced by the camera. This issue can be addressed by using a panoramic camera. Furthermore, panoramic cameras do not require a rotator, which means that the overall system is simpler. However, this system introduces a new problem. With panoramic images taken at a resolution of 2K＠360°, the resolution of the image that the typical user is viewing is actually around 0.5K＠90°, and the bitrate for the resolution is also increased significantly over what had previously been necessary. It might be easier to get your head around it if you think about it this way: when you go from needing to look at only one side of a die, to needing to see all six sides of the die, the bitrate gets six times higher. This is a constraint upon the streaming of 360-degree video (which is 1-to-N, one-way communication), with which the likes of Youtube and Facebook have been grappling. To upload and download high-resolution panoramic video in real-time, a high-performance environment such as that provided by a "PC x Server" is necessary. At the same time, when you combine a "Smartphone x Raspberry Pi", in most cases, you are connecting peer-to-peer (one-to-one, two-way communication), which means that only the image of the one side of the die that the user is looking at needs to be transmitted. Additionally, compared to the "PC x Server" combination, the relative resolution of the smartphone and the relative GPU performance of the Raspberry Pi are offset, and you can expect to have solid performance for the price. For this reason, the Picam360 uses a technology called SSPR(Server Side Panoramic Rendering). We have developed this camera to not only take advantage of the benefits that are gained when you combine "Smartphone x Network x Raspberry Pi x Panoramic Camera", but also to achieve even greater performance by being focused on "Real-Time x Peer-to-Peer".
We originally started work on an open-source hardware project in the hope of making something real out of the idea that if we could mount a camera on the bottom of a boat, people would be able to have the experience of snorkeling from a distance. This would make snorkeling easily accessible even to people who aren't strong swimmers, and in areas that are difficult for people to get to. Putting an emphasis on ease of use, we started by deciding on a 100W solar panel to power the boat. This solves the problem of supplying power to a remote-controlled device, and allows users to have as much fun as they like for as long as the sun is out. The next step was to manufacture a hull, and it proved to be a major challenge, because with open source hardware, it's essential that it can be redesigned inexpensively, and by anyone. Using a 3D printer would have made it possible for anyone to redesign it. However, printing a large object with a 3D printer is expensive and time-consuming. Thus, we had the idea that if we used a 3D printer to print just a mold for the hull, the printing itself would still be a challenge, but the reproduction of the mold would be both less expensive and less time-consuming. We immediately went and set up a Bézier curve in OpenSCAD, and created the framework for a digital hull with only nine control points, which would allow anyone to create a design of their own. With this as the base, we made a digital mold, divided it into sections, printed those sections on a consumer 3D printer, and assembled the hull. With that, the 3D-printed hull mold was a reality. The mold has 90 pieces in total, with each piece taking half a day to print, so it was a pretty big job. As we're sure you can imagine, though, it was a fun experience. By applying a surface treatment, and pouring on some urethane foam, we had a working hull that anyone would be able to design. After that, we chose a propeller, motor, servomotor, and rudders that would work with the 100W solar panel, before designing the necessary parts in OpenSCAD, printing them, and installing them. Accompanied by feelings of hope and trepidation, we took our finished vessel to the lake and tested it on the water. We learned that we could indeed go "remote snorkeling" with this boat. While there were certainly some problems along the way, this phase of the project was another enjoyable one. It's our hope that this design will help others to create new value of their own.
With your support, our ultimate goal for this project is to develop the following features, which will use the functionality of the SurfaceWalker to help solve environmental problems.
A laboratory at the university that has collaborated with me on the development of the SurfaceWalker has launched a project involving the use of inexpensive unmanned vessels to study the quality of lake water. When the quality of the water in a lake decreases, cyanobacteria appear in large quantities. This is a well-known environmental issue, and one that directly impacts human health, as there are some types of cyanobacteria that contain toxins. However, at present, continuous monitoring of cyanobacteria is difficult from a cost standpoint, so we are generally left to take action only after they have already appeared in large numbers. In order to gain a full, up-to-date understanding of this problem, it will be necessary to measure cyanobacteria concentrations over a wide range and a long period of time. Therefore, as one approach, the aforementioned research project was launched. In this project, equipment for measuring cyanobacteria is installed on inexpensive unmanned vessels, which perform autonomously-controlled measurements within a large perimeter. As part of this project, we equipped the SurfaceWalker with GPS and gave it the ability to control itself, so that it could navigate to waypoints that were determined without human input. It has now begun automatically storing data sets that include GPS location information, water temperature, and cyanobacteria concentrations on a database server. As the next step, we are developing OpenCyanoMap, a feature that uses the data collected in this way to create visualizations of cyanobacteria concentrations on OpenStreetMap.
OpenCyanoMap will allow anyone to upload data. Currently, laboratories at several universities in multiple countries that have been supportive of this project are uploading data from their own research on lakes. We are aiming to further expand these efforts, and to create a complete map that shows cyanobacteria concentrations around the world.
While working on making "remote snorkeling" a reality, we met a person who grows coral. That person told me about the hope that fueled their coral-growing work: "In the sea where I would come to play when I was a child, there was lots of coral, and there were a lot of fish, but they're gone now. I want to show my own children the sea, the way it was back then, somehow." Hearing about those hopes, we started to think about whether it would be possible to use the SurfaceWalker to document the current state of coral. We do think it's possible, because if we were to stick with the task of documenting the current situation, we might come up with a way to combat the loss of coral, and we think that we would all be able to teach future generations about what that current situation is. In order to make this happen, in addition to the features that already exist in OpenCyanoMap, it will be necessary to have a feature that would enable the filming of coral from directly above, even as the drone is being shaken by waves. Fortunately, thanks to panoramic cameras and attitude sensors, it is possible to use image processing to extract these vertical images. There are other things that the SurfaceWalker will need in order to be able to do this, such as a feature that will patch together images that have been captured in this way, as well as greater stability on the sea. The most important part of all will be finding people around the world who are willing to help with this project. Supporters will receive a SurfaceWalker in return. We would be delighted if we can all create a record of the current state of coral around the world together.
We're developing a time-lapse feature that will allow you to see how coral has changed by browsing a timeline of collected images. The animation below is a conceptual example that shows a time-lapse of a patch of coral that has been filmed at regular intervals.
When you contribute an image that you have filmed with your Surface Walker to OpenCoralMap, your name will be shown on a credit tag. Credits are shown when you enlarge the map to the necessary size.
All funds that we have received from our supporters will be allocated to cover the expenses listed below.
Risks and challenges
Our greatest concern is our stock of the SurfaceWalker's components, including the cameras and motors. We have secured a supply that will allow us to cover our KickStarter rewards. However, in the unlikely event that there is a supply-related problem, such as our supplier ending production without warning, it is possible that shipments will be delayed so that we can find substitute components and verify their performance. If an unexpected event has occurred, we will provide our backers with clear information in an open manner.Learn about accountability on Kickstarter
- (60 days)