Additional Refinement, Updated Timeline (Trident Kickstarter Update #17)
Original post : December 13, 2016
What to expect in this update:
- Updated Shipping Information
- Design Updates (including a big update to the topside design) & Current Work
- Video from MBARI Testing and Demo and 360 video from Lake Tahoe
Most importantly: we’ve encountered some additional delays - see below for details. Trident pre-orders will ship in the order they are received (your reservation will hold your place in line), starting summer of 2017.
We take schedule very seriously and we want everyone who has supported us to know what we’re doing to make sure they get a product that is extremely well built, but delivered no later than necessary. As we get closer to the ship date, we will create a shipping schedule so that each customer will be able to see which week their unit will ship. Expect that calendar in a coming update.
The good news: the project has not encountered any major problems, just a number of small issues that have added up. Here’s an overview of everything that’s happened and everything that we’ve done:
Prototype shell validation
Before ordering tooling, our procedure has been to go through a series of design and engineering validation tests. This is done both in-house and with various outside experts to ensure quality. As part of this process, we ordered a series of prototypes of the final design. These prototypes were delayed for several weeks at the factory due to errors made by the manufacturer during the build process. This in turn delayed the validation process needed before moving forward with ordering tooling.
We made our original material selection based on depth requirements, but later learned that we couldn’t use the plastic we selected due to certification requirements related to battery and fire protection. Finding a new plastic that met all our criteria took several weeks of research, but we’re excited and confident about what we found. And given everything that has made the news with battery problems lately, we feel like spending extra time and attention in this area is justified.
Navigating through various domestic and international certification requirements for a commercial product also caused delays as a result of required research. For example, there was a period when we believed we might be required to put finger guards over the propellers. After extensive research and verification from experts, we eventually learned that due to the maximum rotational speed and mass of the propellers, guards would not be needed. While this was being confirmed, we had to wait before locking down the shell design. There were numerous other similar instances like this which caused delays in development, but eventually we got to the point where experts said that (pending several tests we can only do once the final production models are made) we should be fully compliant as a consumer product.
Doing in-depth tolerance stackup analysis before preparing the final mechanical drawings revealed several areas of possible concern. Because of the depths to which Trident will go, dimensional tolerancing is critically important as it can affect seal reliability as well as stress distribution throughout the vehicle. Rather than risk problems in tooling, we modified our mechanical drawings to improve the engineering specifications for the design.
Sample part validation
Several of Trident’s smaller parts that we ordered samples of (custom wiring harnesses, fasteners, electrical connectors, adhesive materials, etc) turned out to either not meet specification or needed redesign; this required re-ordering of samples before moving forward. We have a constant and continuous commitment to quality. This rigorous supplier vetting will improve the quality of the product we ship and avoid last-minute production delays.
Motor connector design
Even though we are aiming to make the motors for Trident reliable enough to last the lifetime of the vehicle, we decided it would be worth making them replaceable using basic tools, for cases where something out of the ordinary happens. This design process rippled through other aspects of the vehicle which took several weeks of CAD work to solidify. Our experience from working with other ROVs has taught us that even when things are well designed, moving parts are always something that should be replaceable.
Testing, Testing, Testing...
Over the last few months, we’ve also done a tremendous amount of field and in-lab testing.
Once we received prototypes of our new shell design, one of the first tests we did was an empirical depth rating test in our pressure chamber. We did extensive Finite Element Analysis (FEA) to assure that the vehicle would handle 100m, but testing a physical system in actual water pressure was needed to make sure our calculations were correct. We had done tests like this before with earlier models, but since the new model was our final production design, these series of tests were the ones that mattered.
Our design methodology for motors has been to do extensive testing in harsh conditions and to continuously iterate the design based on our observations of performance and durability. We’ve used delay time to do progressively longer motor tests, and to make additional iterations to the design. So far our test on the newest motor design has been running for more than 400 hours in salt water filled with silt, and there has been no noticeable degradation in performance.
An often overlooked aspect of ROV design is thermal management. Although hot parts of the vehicle can be sinked to water during operation, a good ROV design will allow it to run in air for extended periods of time without failing. Although this may seem trivial, ROVs are sealed containers so they have very little in the way of convection. Aside from convection, the only other ways to dissipate heat are through conduction and radiation. To maximize this cooling capability, we’ve done analysis on the thermal paths in our design and have optimized the system to spread out heat from hot spots on the camera and onboard computer.
Summary: We’re moving away from the towable buoy design towards a smaller, more modular topside WiFi design.
The WiFi Topside Module was one area where we made the decision to save time instead of adding capability. In our original Kickstarter campaign (and also in our update on April 30th) we described the WiFi Topside Module (the part on the other end of the tether that allows you to connect to the vehicle using WiFi) as a buoy that would be shaped in a way that it could be thrown in the water and towed behind Trident.
We built several early prototypes of this design toward the beginning of the project, and were intrigued by the possibilities. But after further testing, we realized that making the WiFi Topside also a towable buoy would add a large amount of risk to the functionality and development of the system as a whole. To keep the buoy from being towed beneath the surface, it would need to be fairly large (nearly the length of Trident itself), to make it self-righting, it would need a large keel ballast which would increase its displacement and add a lot of drag, and finally, to make sure people wouldn’t lose the system, we’d need to do extensive development work with the antenna and radio to assure a solid connection and develop a return-to-home feature if the vehicle lost signal far from shore.
While all of these features could have been designed, they would add months to the development process and could reduce the reliability of the system. We decided to stick with a simpler design that we knew we could trust. The new topside module is something we are very proud of. It is simple and robust, and makes the system a lot more transportable due to its shape and size. Although it is not a floating buoy, it’s still designed to be waterproof and positively buoyant so if it falls overboard accidentally, the system won’t be lost. Though not hydrodynamic, users on the water who (at their own risk) want to decouple the buoy from their boat may still be able to experiment with putting the buoy on other floating devices. A wifi buoy is just one of the possibilities for the modular, smaller design. We’ve talked with a few people about the new design, and have already heard several ideas of how folks will be adapting it to their specific purpose. We’ve set up a thread on our forums to discuss ideas.
A big perk of the new Topside design is that it is now small enough to be attached to the hub of a tether reel, which will make tether management much more elegant for people who ordered a 100m length of tether. The image below shows a 3D printed prototype of the WiFi Topside Module (we’re still waiting for the nicer prototypes to come in) attached to a tether reel. This configuration allows the reel to be used without a slip-ring because the WiFi Topside can just rotate with the tether.
We’ll post photos of the professionally-built WiFi Topside Module prototypes on the forum thread when they come in, but in the meantime, here are some renderings of the design.
Because the smaller Topside Module will now fit well in a reel, we’ve begun to shop around for simple reel systems to include with the 100m tether accessory option. The 25m length that comes with Trident is short enough to coil hand over hand easily, but our work in the field has taught us that 100m can be a bit long for that. The tether reel will be basic, but we hope it will make management of the longer length much easier.
Earlier this week, we had the opportunity to demo and test Trident at the Monterey Bay Aquarium Research Institute (MBARI) in Moss Landing, CA. From the beginning, we wanted to make sure Trident was more than just a camera in the water. It had to stand up to rigorous scientific use. Demoing for the team at MBARI gave us the opportunity to put the device in the hands of people who are on the leading edge of ROV and AUV development and operation.
Here is a link to the Facebook Live post we made while talking with the Director of Engineering at MBARI, Doug Au, while on site:
360 Camera Field Testing in Tahoe
A number of backers have asked about attaching 360 cameras so we wanted to test and demo how that would work using Trident’s payload mounts. Here’s a photo of the new waterproof Nikon KeyMission 360 mounted to Trident:
We took it for a test run in Lake Tahoe (about three hours east of our lab). We chose to go to the site of a small sunken sailboat about 50m from shore in approximately 20m of water. We used a second Trident to film the first. The resulting footage is a lot of fun to watch!
We’re still going through footage from both Tridents’ onboard cameras. In the coming days, we’ll try to edit that footage down and post a highlight reel on our YouTube channel.
We’ve been ramping up our field testing program now that prototypes of the final design are arriving. Many people have said they’d love to see more videos of Trident flying, so we are hoping to have a lot more of that to show off soon. If you have ideas about places to test, or you’d like to take part in some of these testing expeditions, please let us know in the comments below. We appreciate the patience everyone has had as the project comes together, and we continue to strive toward making the best product we possibly can.
As always, feel free to reach out directly with any questions, concerns, or ideas.