Mechanical
Subsystem

01

System Overview

The initial phase of the JellyBot project successfully validated the feasibility of pneumatic biomimetic propulsion using a soft-robotic silicone bell. While Prototype 1 demonstrated the core swimming mechanics, the interim testing identified critical "pain points" regarding waterproofing reliability, material fatigue in the pneumatic lines, and the lack of internal structural organization.

Phase II of the mechanical development focuses on professionalization and ruggedization. The primary objective was to transition from a proof-of-concept PVC housing to an industrial-grade underwater enclosure system. This semester’s work involved the mechanical integration of a Blue Robotics cylindrical housing, the fabrication of custom marine-grade cable penetrators, and the optimization of the silicone casting process to ensure uniform propulsion. These enhancements aim to move JellyBot from a lab-based prototype to a functional, deployment-ready surveillance platform.

02

Summary of Preliminary Development (Phase I)

The first phase of the Jellybot project focused on establishing a baseline for biomimetic underwater propulsion. The primary goal was to validate the “Soft Pneumatic Actuator” concept and integrate it into a waterproof housing capable of basic submerged operation

2.1 Concept Selection and Prototype 1 Design

Following an evaluation of various propulsion methods, Concept C: Pneumatic Silicone Actuators was selected for its high biomimicry and low ecological impact. Prototype 1 was constructed using a cylindrical PVC pipe as the main structural body and a clear acrylic dome for the camera housing. The design prioritized cost-effectiveness and rapid iterative testing over industrial-grade depth ratings.

The propulsion system consisted of eight custom-cast silicone tentacles (a 50/50 blend of Ecoflex 00-30 and Dragon Skin 20) unified by a 2mm thick silicone flap. A mechanical flap clap was developed to interface these soft components with the rigid PVC housing, ensuring the assembly could withstand the internal air pressure changes required for pulsing movement

2.2 Key Technical Findings (Phase I "Pain Points")

The lessons learned from Prototype 1 formed the critical requirements for the Phase 2 development. The four primary areas identified for improvement were:

03

Mechanical Evolution & System Upgrades

Based on the technical shortcomings identified in Prototype 1, Phase II focused on a transition from a "Proof of Concept" (PoC) to a "Functional Prototype." This evolution involved shifting from consumer-grade materials to industrial-grade marine components and professionalizing all mechanical interfaces.

3.1 Design Requirements for Prototype 2

To address the "Pain Points" from Phase I, the following mechanical requirements were established to ensure the JellyBot could withstand real-world underwater challenges:

• Repeatability: The housing must allow for repeated opening and closing for maintenance without compromising the waterproof seal.
• Structural Stability: Internal electronics must remain fixed relative to the housing to prevent shifting of the Center of Gravity during swimming.
• Hermetic Sealing: All cable entries must be sealed against hydrostatic pressure using mechanical compression methods rather than temporary surface sealants.

3.2 Comparative Analysis of System Architecture

The following table maps the specific mechanical failures of the interim phase to the engineering solutions implemented in Prototype 2. This transition focused on moving away from manual, additive sealing toward integrated mechanical seals.

Table 1: Comparative Mechanical Specifications and Evolutions

3.2.1 Transition to Industrial-Grade Housing

The move from manual PVC to a Blue Robotics enclosure provided the precision tolerances required for deep-sea reliability. The new cast acrylic and aluminum housing utilizes dual-radial O-ring flanges, which provide a repeatable, high-pressure seal that does not degrade after repeated assembly.

3.2.2 High-Pressure Sealing and Modularity

To resolve the leakage issues from Phase I, makeshift glands were replaced with Mechanical Compression Penetrators. This upgrade eliminates the "manual sealing back-and-forth" and allows for a modular system where cables can be adjusted without damaging the pressure hull.

As seen from the Figure 2, the makeshift glands and silicone sealants were replaced with Mechanical Compression Penetrators. These utilize a specific Head, Shoulder, and Throat geometry to compress the cable jacket, creating a hermetic seal.

3.2.3 Material Homogeneity in Pneumatics

A critical failure in Prototype 1 was the air leak caused by bonding dissimilar materials (plastic tubing to silicone tentacles). In Prototype 2, the system was redesigned to be entirely silicone-based. By using chemically identical materials for both the tubes and the actuators, the silicone adhesive creates a cohesive molecular bond. This "cold weld" ensures that the pneumatic interface remains airtight even under the high internal pressures required for forceful bell contractions.

3.2.4 Internal Structural Organization

To ensure electrical stability, a Modular Internal Mounting Tray was designed and fabricated. In Prototype 1, loose components shifted during movement, risking wire fatigue and changing the robot's Center of Gravity (CoG). The new rigid framework secures the pumps, batteries, and controllers in a fixed orientation, ensuring a stable swimming gait and simplifying the troubleshooting process during field testing.

04

Industrial-Grade Housing Integration

The transition to a Blue Robotics 4-inch Series Enclosure represented a critical move from experimental Proof-of-Concept (PoC) materials to industry-standard marine hardware. This section details the selection rationale and the mechanical design of the components required to interface with this new pressure hull.

4.1 Selection of Outsourced Housing

The selection of the Blue Robotics platform was a strategic decision driven by risk mitigation and the need for a pre-verified underwater environment.

4.1.1 Risk Mitigation and R&D Efficiency

By outsourcing to a pre-verified industrial solution, the team avoided extensive independent hydrostatic testing. Market research revealed that designing a custom pressure vessel in-house would have consumed a significant portion of the semester, detracting from the focus on complex soft-actuator interfaces.

4.1.2 Dimensional Continuity

The team intentionally selected a configuration that mirrored the height and diameter of the Phase I PVC prototype. This ensures that hydrodynamic drag calculations and buoyancy baseline data remain relevant for a direct "apples-to-apples" performance comparison.

4.1.3 Material Selection (Cast Acrylic)

The team opted for Cast Acrylic over Aluminum for three primary reasons:

Blue Robotics provides two primary material options for the cylindrical housing: 6061 Aluminum and Cast Acrylic. While Aluminum offers higher depth ratings, the team opted for the Cast Acrylic version to prioritize the specific surveillance and testing requirements of the JellyBot project.

Table 2: Comparative Material Selection

4.2 Design and Integration of Prototype 2

The design of Prototype 2 focused on the seamless integration of successful biomimetic actuators from Phase I with the professional pressure hull. By maintaining physical dimensions, the team ensured fluid displacement models remained consistent.

4.2.1 Housing CAD

Before designing any internal or external attachments, the Blue Robotics 4-inch Series Enclosure was modeled in a 3D CAD environment to serve as the foundational structural datum for the entire project. This digital representation provided a high-precision fixed reference that informed every subsequent design decision, ensuring that all custom-fabricated components would interface seamlessly with the industrial hardware.

By accurately modeling the 4-inch internal diameter of the cast acrylic tube, the team was able to establish strict volumetric constraints, which were essential for sizing the modular internal mounting tray to achieve a perfect "sliding fit" without mechanical interference. Furthermore, importing the specific geometry of the aluminum end-caps allowed for the precise mapping of the O-ring grooves and pre-drilled penetrator ports, guaranteeing that the sealing interfaces remained uncompromised. Finally, the CAD model served as a critical tool for optical verification; by simulating the camera’s placement within the transparent housing, the team confirmed that the Field of View (FOV) would remain entirely unobstructed by the internal structural rails or the complex network of pneumatic tubing required for actuation.

4.2.2 Designing of Tentacles

The internal geometry of the pneumatic tentacles was reused to maintain the validated bending moment. However, the inlet ports were modified to accommodate silicone tubing, allowing for the cohesive molecular bonding required for high-pressure airtight integrity.

4.2.3 Designing of Flap Clamp 2.0

The Flap Clamp was redesigned to serve as the mechanical bridge between the soft "bell" and the rigid housing without requiring permanent, destructive modifications to the end-caps.

• Friction-Based "Sandwich" Mechanism: A two-part interlocking ring design compresses the silicone flap against the exterior wall of the housing.
• Non-Invasive Mounting: This compression-fit approach preserves the structural integrity of the pressure hull and prevents potential leak paths through additional bolt holes.
• Ease of Adjustment: Allows the propulsion module to be rotated or shifted to fine-tune the Center of Gravity (CoG) and thrust alignment without tools.

4.2.4 Designing of Modular Mounting Tray

To resolve the disorganized layout of Phase I, a custom Modular Mounting Tray was designed as the central structural backbone for the internal electronics.

4.2.5 Integrated CAD System Assembly

The Integrated CAD System Assembly served as the final digital validation phase, where all custom-designed sub-components were virtually mated to the primary Blue Robotics 4-inch Series housing. The primary objective of this stage was to conduct a comprehensive spatial fit check to ensure that the internal mounting tray, the eight-tentacle actuator array, and the 3D-printed flap clamp functioned together as a cohesive unit.

By visualizing the "Digital Twin" of the Prototype 2 design, the team was able to confirm that the internal components cleared the O-ring sealing zones and that the external clamp provided the necessary 360° coverage without interfering with the end-cap hardware. This pre-fabrication review allowed the team to verify all mechanical clearances and spatial orientations in a virtual environment, providing the confidence required to proceed with the physical fabrication and 3D printing of the final components.

05

Advanced Fabrication: Prototype 2 Components

This section details the physical assembly and specialized fabrication techniques used to realize the Prototype 2 design. The focus was on ensuring that every mechanical interface was robust enough to handle the pressures and vibrations of an underwater environment.

5.1 Installation and Integration of Cable Penetrators

To resolve the leakage and "wicking" issues identified in Phase I, Blue Robotics mechanical compression penetrators were integrated into the aluminum end-caps.

5.1.1 Installation of Electrical Cable Penetrators

For the power and data tether, standard Blue Robotics Mechanical Compression Penetrators were used.

5.1.2 Custom Modification: Pneumatic "Bulkhead" Penetrators

A significant engineering challenge in Prototype 2 was the management of air supply. Blue Robotics does not manufacture a dedicated pneumatic penetrator, and standard cable penetrators cannot facilitate the high-flow air required for soft actuation.

• The Innovation: The team developed a custom hybrid solution by modifying an industrial-grade Pneumatic Barbed Bulkhead Connector.
• The Process: An industrial pneumatic coupling was precision-fitted into the penetrator throat. The silicone pneumatic tube was treated as the "cable," locked securely by the internal compression mechanism.
• Result: This innovation provided a robust, watertight air-path, using the reliable O-ring sealing of the Blue Robotics system to facilitate high-pressure pneumatic actuation in a submerged environment.

5.2 Optimized Silicone Casting (Tentacles & Flap)

The fabrication process for the soft actuators was overhauled to prioritize chemical bonding and structural uniformity, resolving previous inconsistent propulsion issues.

5.2.1 Material Homogeneity and Bonding

A critical technical upgrade in Prototype 2 was the transition to a fully silicone-based pneumatic circuit. In the previous iteration, the use of dissimilar materials, specifically plastic tubing and silicone tentacles, prevented a secure chemical bond, leading to frequent air leaks at the high-pressure inlet.

By switching to silicone tubing, the team matched the material properties of the tentacles. When a specialized silicone adhesive is applied, it creates a "cohesive bond" rather than a mere surface-level attachment. Because both the tube and the actuator share an identical chemical base, the adhesive effectively fuses them into a single, airtight unit. This "cold weld" effect has significantly increased the system's pressure tolerance during the high-frequency pulsing cycles required for propulsion.

5.2.2 Standardized Casting Formula and Preparation

To ensure uniform thrust across all eight actuators, the team implemented a rigorous preparation protocol to eliminate the material inconsistencies that plagued Prototype 1. While the core casting formula remained the same, the transition from estimation to precision measurement was a key driver in achieving structural uniformity.

• Precision Measurement: Unlike the "by-eye" estimation used in the previous phase, a precision digital scale was used to maintain a strict 1:1 mixing ratio of Smooth-On Ecoflex 00-30 and Dragon Skin 20. This eliminated variations in shore hardness between tentacles, ensuring each actuator possessed identical elasticity and bending characteristics.
• Improved Mold Sealing: To address the "leakage" issues during the curing process, the original Phase I molds were reinforced with taping around the interface seams. This manual sealing technique ensured that the silicone remained contained within the mold cavities, preventing the formation of "flashing" or voids. This resulted in more accurate tentacle geometry and significantly reduced the structural informality that previously compromised propulsion symmetry.

5.3 Internal Tray Assembly

The internal structural backbone consists of vertical aluminum rails and 3D-printed mounting plates, providing a rigid framework that slides directly into the acrylic tube.

Table 3: Modular Tray Benefits

5.4 Fabrication of Flap Clamp 2.0

The Flap Clamp was 3D-printed using PLA for its high rigidity. As the stainless steel bolts are tightened, the PLA rings squeeze the silicone against the acrylic tube, creating a high-friction "clamp-down" effect that anchors the module without permanent adhesives.

5.5 Fabrication of Ruggedized Ground Control Station (GCS)

To transition JellyBot from lab-based testing to field operations, a custom Portable Ground Control Station (GCS) was developed. This unit serves as the primary operator interface, allowing for real-time surveillance monitoring and robot control in wet or high-glare environments where a standard laptop would be impractical.

5.5.1 Structural Housing and Cost Effective Design

• Foam-Base Integration: To minimize fabrication costs while maintaining structural integrity, the team utilized a high-density Pick-and-Pluck foam as the primary internal chassis. The foam was custom-contoured to cradle the electronic components, providing a lightweight yet secure mounting method that protects the hardware from mechanical shock during transport.
• Power Management: An external toggle switch was installed on the side of the case, allowing for rapid power-up. The unit is fully autonomous, powered by an internal 12V rechargeable battery, eliminating the need for external AC power cords near the water's edge.

5.5.2 System Architecture and Interface

06

Verification: Phase II Waterproofing Results

To validate the mechanical integrity of the Prototype 2 upgrades, a rigorous series of waterproofing and pressure tests were conducted. The objective was to confirm that the professional housing and custom-modified penetrators could maintain a hermetic seal under hydrostatic pressure.

6.1 Verification Methodology

The testing followed a "bottom-up" approach, where individual seals were verified before the entire system was submerged:

• Indicator Material: Dry tissue paper was placed inside the housing as a highly sensitive visual indicator for water ingress.
• Leak Detection: Water was mixed with Red Dye to ensure that even a microscopic leak would be immediately visible as a stain on the internal tissue paper.
• Success Criteria: A "Pass" was defined as a 100% dry interior with zero dye staining after a minimum 10-minute submersion at a depth of 1 meter.

6.2 Phase I: Sub-Assembly Housing Tests

Before installing any electronics, the primary structural seals of the Blue Robotics enclosure were tested. The acrylic dome and the aluminum end-cap were tested independently for 10 minutes. By using Red Dye in the water and dry tissue paper inside, the team confirmed that the dual-radial O-rings provided a perfect seal.

Result: No dye ingress or moisture was detected on the tissue indicators.

6.3 Phase II: Electrical Cable Penetrator Verification

The next stage involved testing the mechanical compression of the electrical tether interface. The cable was stripped and installed into the penetrator as per Section 5.1.1. The assembly was submerged at 1 meter.

Objective: To ensure that the "Shoulder" and "Throat" geometry of the penetrator effectively compressed the cable jacket to prevent water "wicking."

Result: The internal wiring remained 100% dry, validating the transition from makeshift glands to professional penetrators.

6.4 Phase III: Custom Pneumatic Penetrator & Pressure Testing

Because the pneumatic penetrator was a custom modification (Section 5.1.2), it required an additional Positive Pressure Test before submersion. During the first air-pressure test, minor bubbling was observed at the threaded interface of the industrial barbed coupling.

Corrective Action: Teflon Tape was applied to the threads to ensure a gas-tight seal. Once the air-leaks were resolved, the pneumatic penetrator was submerged at 1 meter. The combination of the Blue Robotics bolt and the modified coupling successfully blocked all water ingress.

6.5 Phase IV: Final Integrated Prototype Test (1.0m Depth)

The final stage of verification involved submerging the fully assembled JellyBot to simulate a real-world surveillance mission. Unlike the previous sub-assembly tests, this phase included the installation of the Internal Mounting Tray, the pneumatic pumps, and the control electronics.

The robot was fully sealed with the 3D-printed Flap Clamp and soft bell attached. The internal tray was populated with all components to ensure the Center of Gravity (CoG) was realistic, and the tether was connected to the Ground Control Station (GCS). The entire prototype was submerged in a test tank at a depth of 1.0 meter for a duration of 10 minutes.

Table 4: Summary of Final Integration Results

6.6 Final Mechanical Readiness

With the successful completion of the integrated full-system submersion test, the JellyBot Prototype 2 has successfully met all "Phase II" objectives. This iteration represents a definitive technical leap over the previous version, systematically resolving every "Pain Point" identified during the interim phase:

The Prototype 2 architecture is now mechanically validated and proven to be hermetically sound. Having withstood the rigors of submerged pressure testing, the system has demonstrated the durability required to overcome environmental challenges. The team is fully confident in the robot’s structural integrity, marking JellyBot’s readiness to transition from controlled laboratory environments to real-world, long-duration coral surveillance missions.

07

Future Work: Operational and Hardware Enhancements

While Prototype 2 has achieved full mechanical validation, several hardware and operational optimizations have been identified to improve the JellyBot’s efficiency and mission capabilities in real-world marine environments.

7.1 Integrated External Power Switching

A primary operational limitation of the current prototype is the requirement to partially disassemble the pressure hull to manually connect or disconnect the internal battery. This process is time-consuming and increases the risk of introducing humidity or debris into the sealed electronics chamber, which can lead to long-term component degradation.

Proposed Solution: Future iterations will resolve this by integrating a waterproof External Piezo Switch directly into the aluminum end-cap. By adopting this industry-standard approach, the JellyBot can utilize an M10-threaded penetrator switch to maintain a hermetic seal while allowing for external power control (Blue Robotics, n.d.). This modification would allow the operator to power the Raspberry Pi and the pneumatic systems on or off from the exterior of the housing.

7.2 Integrated Ballasting and Weight Distribution

Currently, the prototype relies on external lead weights that are manually suspended from the exterior of the JellyBot. This manual setup is inefficient, as it requires constant re-attachment and complicates the deployment workflow.

Proposed Solution: Future iterations will integrate a structural ballast. It will utilize the Flap Clamp as a multi-functional component. Since the specific buoyancy requirements have been characterized during Phase II testing, we can transition to a weighted-core fabrication method. We can fabricate a replacement Flap Clamp designed with the exact mass required to achieve neutral buoyancy. By utilizing high-density inserts or outsourcing production to include a custom-weighted core, the ballast becomes a permanent part of the robot’s structural assembly.

Appendix

List of Figures

Appendix

Tables List

Appendix

References