The Cotton Candy Enigma: Exploring Material Deposition Through Miniature Mobile Robots and Augmented Reality

The Cotton Candy Enigma revolves around three core concepts: leveraging a cotton candy machine mechanism, integrating an augmented reality (AR) application, and exploring the potential of compact mobile robots to develop a workflow that uniquely generates temporary enclosure. The cotton candy machine serves as the fundamental basis for this project, influenced by several scholarly papers discovered during the research phase. The intricate mechanics of the machine enable a distinctive method of aggregating and depositing materials, such as sugar, with the possibility of future adaptation to various plastics like PLA.

Our objective was to combine the cotton candy machine with mobile robotics, thereby establishing a continuous feedback loop encompassing environmental scanning, material deposition, automated refilling, and spatial creation. The primary focus of this endeavor involved the abstract representation and miniaturization of a machinery component.

The creation of this project can be attributed to the collaborative efforts of Hamed Bemanesh, Cornelius Carl, and Sam Losi, who embarked upon its development during the Computational Design and Digital Fabrication course at the University of Stuttgart.

The majority of the structural components and connections were fabricated using 3D printing technology. This approach offered several advantages, including the ability to rapidly prototype and iterate designs, as well as the flexibility to customize and optimize the parts for their intended functionality. The utilization of 3D printing minimized material waste and enabled the creation of intricate and precise geometries, ensuring a robust and lightweight construction. The rest of the materials are listed here:

(1) raspberry pi 4

(1) CPU fan 120mm

(1) DC motor

(2) Cartridge Heaters

(2) (10mm×10mm) Aluminium sheet

(2) Servo motor

(2) Wheel

(A few) M2 bolt

(A few) Wires

The initial phase of this project involved the abstraction of the mechanical aspects of the cotton candy machine. To gain a deeper understanding of its functionality, our team acquired a second-hand cotton candy machine and proceeded to deconstruct and test it. Through these initial tests, we not only enjoyed some delectable treats but also gained valuable insights into the inner workings of the machine.

Fundamentally, the machine operates by employing two key techniques: centrifugal force and heating. At its core, the machine features a central geometry known as the sugar chamber, where the user places the sugar. Upon activating the machine, this chamber rapidly spins with great velocity. The chamber incorporates local radiant heating elements, which generate heat that is transferred to the sugar chamber walls. Consequently, the sugar transforms into a molten state due to this heat, and the centrifugal force exerted by the spinning action drives the molten sugar towards the edges of the chamber.

To facilitate the process, the edge of the chamber includes a gap that restricts the passage of solid sugar but allows the liquid sugar to flow through. As a result, the spinning force propels tiny liquid strands of sugar in all directions, ultimately into the collection bowl. Upon contact with the cooler surroundings, these thin liquid strands rapidly solidify into delicate, thread-like structures. We have theorized that with the ability to manipulate and direct these strands, it becomes feasible to aggregate them onto a scaffold, thereby creating volumetric forms.

Subsequently, a series of meticulous tests were conducted to determine the optimal approach for directing the deposition of sugar fibers. Initially, our team drew inspiration from previous research papers and considered utilizing an air compressor. To implement this approach, we created a circular tube with punctured holes, which was connected to our test device. By regulating the airflow from the compressor, we attempted to achieve desired results. However, the initial tests, which we have provided video documentation for your reference, did not yield the desired outcome. We discovered that the air compressor, designed to deliver a substantial volume of airflow, generated excessive force in specific localized areas. Given the lightweight nature of the liquid sugar strands, it became evident that a different method of air movement, capable of handling a larger volume of air even at a slower pace, would be more suitable.

Consequently, in the second round of tests, also documented in the attached videos, we positioned a standard desk fan directly behind the cotton candy machine, tilted at approximately 30 degrees. These tests surpassed our expectations in terms of success. Notably, we made two significant observations. Firstly, when the surrounding air was in uniform motion, the spinning motion of the sugar chamber created a vortex-like effect that entrapped the surrounding air. This resulted in the sugar strands being drawn into the vortex and shaping themselves into tubular forms. As the accumulation of these strands progressed, they began to be propelled in the direction of the airflow. The strands were successfully captured onto a scaffold constructed from sticks, which we believe could be further optimized to enhance the efficiency of sugar strand capture. These findings significantly reinforced the potential of this material deposition method, prompting us to shift our focus towards the miniaturization of the mechanism and recognizing the genuine promise it held.

Following the tests, it was determined that separate hardware components for spinning and heating would be necessary. For the spinning mechanism, we opted to use mini heating elements akin to those commonly found in 3D printers, alongside a DC motor. However, a major hurdle emerged when considering the need for the sugar chamber to rotate while maintaining the transfer of electricity to the heating elements.

To overcome this challenge, we disassembled a second-hand vacuum cleaner upon realizing that it housed the precise mechanism we required: a Slip Ring. This component enables the transmission of electricity from a static element to a rotating one. The Slip Ring consists of two integral parts: spinning copper rings known as brushes and a static copper piece referred to as shorting bars. Throughout the entire rotation of the device, these components remain in continuous contact, facilitating the uninterrupted transfer of electrical power.

The miniature airflow system was also designed and integrated. We envisioned utilizing multiple CPU fans strategically positioned to capture and direct the airflow in accordance with the design of the miniature sugar chamber. This approach was aimed at ensuring efficient and controlled movement of air around the chamber, facilitating the formation and deposition of the sugar strands.

Armed with this newfound knowledge, we proceeded to design and prepare the necessary specifications for the creation of a miniature cotton candy mechanism, incorporating the Slip Ring and the other identified hardware components.

In the design of the mini mechanism, the geometry of the sugar chamber played a crucial role. To simplify the mechanism, we opted for two curved opposing plates. Initially, we engaged in discussions and attempted to shape store-bought metal cans, but these efforts proved unfruitful. Eventually, we determined that aluminum would be the ideal material for the plates, benefiting from its excellent heat conduction and lightweight properties.

Although the preferred approach would have involved iterative machining of these parts, the project timeline imposed limitations. As a result, we needed to explore an alternative solution that was quicker and more manual in nature. We arrived at the idea of creating two components of a 3D printed mold, with a 100% infill for enhanced strength. These mold pieces could be positioned on either side of a metal sheet and pressed together using a vice. Numerous iterations of this process were conducted, leading us to realize that the molds required an additional flat, mold-like piece around their periphery to prevent the surrounding metal from creasing or bending.

By employing this methodology, we successfully achieved the desired shape for the curved plates while adhering to the project's timeline constraints.

The assembly of the miniature mechanism, similar to the project as a whole, unfolded through a series of iterative steps. Several noteworthy events occurred during this process. Firstly, we encountered tolerance issues concerning the heating elements and the assembly of various components. This led to a situation where, upon turning on the motor, instead of gently forming the sugar strands, the mechanism forcefully propelled the heating elements across the room. Consequently, we recognized the importance of safety precautions and made it a practice to wear protective glasses during subsequent assembly stages.🙂

Additionally, as we progressed with the assembly, we realized that the power of our initial CPU fan design was insufficient to generate the desired airflow. In response, we were able to acquire a single, more powerful fan that surpassed the combined output of the smaller fans we initially considered. Ultimately, we positioned the miniaturized mechanism at the center of this larger fan, carefully routing the wiring to maintain a predetermined path.

These experiences highlighted the importance of adaptability and problem-solving throughout the assembly phase. We continually evaluated and adjusted our approach to address challenges as they arose, ensuring the functionality and effectiveness of the miniature cotton candy mechanism.

Please see the attached photos for the assembled mechanism, a tangible glimpse into the realization of our vision and serve as a testament to the successful integration of the various elements discussed throughout our discourse.

The development of the accompanying mobile application involved utilizing Swift programming language, along with SwiftUI framework for the user interface design. Additionally, we leveraged the capabilities of RealityKit and ARKit frameworks to facilitate the scanning of the environment and enable the precise placement of the robot within the augmented reality (AR) space.

The user interface (UI) of the app was thoughtfully designed to guide users through the scanning process seamlessly. It provides clear instructions and prompts to assist users in capturing the necessary environmental data for accurate placement of the robot. Furthermore, the app offers intuitive controls for users to position and orient the robot within the AR environment, ensuring precise alignment with their desired location. Notably, the interface also provides visual indicators highlighting the locations of the struts, aiding users in achieving optimal positioning and alignment.

Upon successful placement, the app requires users to connect to the Wi-Fi network generated by the Raspberry Pi embedded within the robot. This connection enables the seamless transmission of movement commands from the app to the robot. Additionally, the app monitors and displays real-time data from various sensors on the robot, including speed, temperature of the spindle, and the robot's velocity. This comprehensive monitoring capability allows users to have full awareness and control over the robot's performance during operation.

To enhance the mobility of the robot within the environment, we incorporated wheels and calibrated them based on the movement paths generated by the AR app. This calibration process ensured precise and accurate navigation, allowing the robot to move freely and explore areas beyond the reach of an industrial arm robot. In order to establish wireless communication for the mobile robot, we employed a Raspberry Pi, which integrated a Wi-Fi module. This setup enabled seamless communication and control between the robot and the AR app.

To achieve continuous movement, we utilized modified servo motors that were centrally housed within the robot body. These servo motors enabled smooth and precise motion, contributing to the robot's overall agility and maneuverability. Additionally, we strategically positioned two operational wheels at the back of the robot, ensuring unimpeded movement even when the material deposition process was in progress. At the front, we incorporated an inoperable ball wheel to provide stability and facilitate smooth navigation in conjunction with the other wheels.

In conclusion, the arrangement of these components on the mobile robot represents a larger system where multiple robots operate as a fleet, collectively shaping the space through their unique method of material deposition. This coordinated effort not only showcases the individual robot's capabilities but also highlights the potential for a collaborative and synchronized approach in shaping and transforming the environment.