Why am I doing this?
A part of this project is defining the real world niche this project fits in. Every year, thousands of kilograms of illicit substances are smuggled into the country via small, disposable submarines. These subs carry millions of dollars worth of material, and are very hard to detect. In order to improve the safety and security of the US, we had to design an autonomous boat platform.
The Big Picture
The goal of this project was to design a system from the ground up, tying together electronics, communications, as well as design and placement. Practically, everything we designed from scratch.
Thought process
From my time sailing, I knew a catamaran design or a tri-maran would function the best. In addition, due to the Portsmouth Yardstick racing handicap & America’s Cup Racing, I knew the LOA of a boat generally determines its speed. From my familiarity, we opted to go with a scaled down version of a Hobie 16 Racing Catamaran. We opted for this because they are stable, fast catamarans with low displacement, capable of carrying high loads while still being very fast. In addition, the shape of the hulls causes them to track straight without rudder intervention, as well as enables the boat to slice through waves, rather than bobbing up and down on them. Practically, this process took a well known, tested system, and implemented it in a slightly different way. In terms of the control schema, a traditional steering wheel was used with a 9DOF system in order to determine the rotation of the wheel.
Background information
In terms of boat design, significant amounts of detail with shaping the hulls were required. In addition, waterproofing the electronics and mounting hardware was a key component.
In regards to control systems, the Xbee 2.4 Ghz series 1 was implemented. In combination with two arduinos, relay shields, and a 9DOF sensor. These sensors and their accompany libraries, took a 3d printed wheel and turned it into a steer by wire control system.
Design process
Two stages of design happened
The boat:
The hulls of the boat were constructed using foam and sealed with mod podge and painted pink! This required a lot of sanding, cutting, and most importantly, precision as irregularities in the hull would have caused more issues.
The control:
The control system of the boat required additional sensors, allowing it to transform from the basic rotary potentiometer, to a wheel capable of rotation. This simple transformation made it dead easy to steer. However significant amounts of software had to be custom written to process the remote control data as well as a custom message protocol was developed to send the proper heading adjustments.
Results & Discussion
The boat worked great, it was definitely the fastest on the water. On the other hand, the servo motor locked, and would continually jam. The drive by wire control system was definitely very intuitive, however it was hamstrung by the arduino uno that the team needed to lug around. With a bit more time, we could have significantly miniaturized the system. In addition, the hulls and rudder functioned as intended, yielding fantastically responsive turning. All in all, upgrading the servo would have won us the race, and I am sad that we didn’t take the time to do that.
Looking back at my focus, I wish I had time to implement a bearing steering system akin to what was on the controller already. The system would have enabled us to lock in a heading and stay on the heading. In addition, I wanted time to air wire the system as the breadboarded components were prone to increased sensitivity. Finally, miniaturizing the system should have been possible with an additional day to make everything fit on the wheel.
Fluid Sim of the Hulls
Look at how nice these Look!
The Fluid Sim of the Hulls