Learning Experience

Design constraints

A recurring problem we faced throughout the project was dealing with design constraints and how important they are for portability.

In the beginning, the only constraint we had to deal with was the weight which could be supported by the balloon. Our choices were mostly driven by how much each component to be added would affect the final weight of the Hoverfish. Moving on, however, we realised that we had gone overboard with our single-minded focus on weight, driving down wriggle room we had for other constraints such as power and the torque of the servos. Furthermore, we were also hit by this realisation on the software side of things, with the Pi hitting a 100% CPU load at times during our testing.

As we were faced by a greater number of constraints, the name of the game changed to min-maxing various design elements while trying to meet the original design objectives of the project.

The value of experience

Tied in to the previous point, through failing a good number of times, we were reminded time and time again that our estimates based upon a lack of experience were far from accurate. Despite looking through many articles, videos and even papers on estimating the lift of a balloon, nothing beat filling up our balloons and feeling their lift for ourselves. Not only was it dependent on the size of the balloon, we discovered that the balloon material too had a large impact on its carrying capacity. Statistics on torque provided and power requirements were nothing compared to having the Pi restart repeatedly due to a lack of power or our servos failing to lift their load.

What we discovered from our experience was that estimates, even with pre-calculations done can be far off from the final product, due to factors such as craftsmanship and inaccuracies in building. While we tried to cover this by providing ample space for correction, one needs to have backup plans when things turn out better, or worse than expected.

Working with mylar

Mostlly covered in the development-blimp section, we went through 6 prototypes, discovering the properties of different types of mylar, the method of sealing the envelope and how to go from design to the inflated envelope.

Touch controls with Javascript

Covered in the Pi development, this was a fun web-based development skill to pick up, which opens up many possibilities for interacting with touch sensitive devices.

Working with Fusion360

Fusion360 is a powerful 3d modelling tool that preserves your sketches and model extrusion in a timeline series. In this project, we often find ourselves having to go back and change the dimension of the 2d sketches so that it fits with our parts perfectly. Fusion360 will dynamically change all the future work when we alter any changes in the past, making it a very powerful tool in rapid prototyping of our 3d printed components