P17227: Walking Soft Robot

Preliminary Detailed Design

Table of Contents

Team Vision for Preliminary Detailed Design Phase

Phase Summary:
Phase Plan:

At the beginning of this phase, the main objective was to continue development with researching various materials and adhesives necessary to support and move the robot. Through prototyping, analysis, and research, concepts were more solidified and the design progressed. The method of walking was also investigated between multiple disciplines to verify the feasibility among all involved parties. Electrically, further research was planned to be completed and the comprehension of code was to be realized. During this phase, hardware test plans were initiated and components were allocated to perform the various tests listed below.


We continued development and solidification of our material concept. We were also able to create a more detailed design of system integration necessary to support and move the robot. This was done through additional prototyping, analysis, and research. More detailed test plans were developed during this phase in order to carry out more precise testing for next phase.

Feasibility: Prototyping, Analysis, Simulation


Bone Version 1

Version 1 of the bone consists of a simple rod with rounded ends. The rounded ends were meant for articulation between two bones in the likeness of a joint. Gluing the edges in this fashion proved to be impractical, and the rounded ends were not ideal for a joint. A simple hinge was more ideal and more simple to make and use.

Bone Version 2

Version 2 was a simple rectangular rod with square ends. Separate pieces of material are used to create extensions that act as the pivot points for the hinge design. The rectangular body was made from a single folded piece of fabric, which proved to be the issue with this design. Due to the thickness of the material, it was difficult to fold and glue to single sheet of fabric in the small and detailed design.

Bone Version 3

Version 3 was designed to address the manufacturing difficulty of version 2. The rectangular rod was split into a shaft and two end caps. Additionally, simple wooden blocks were made to assist in the molding and gluing of the fabric. Version 3 retains the separate material for the hinges identical to version 2. The changes to the end caps is the focus of the image above.

Soft Robotic Muscle Version 1

public/Photo Gallery/Old Actuator.PNG

The first model of the soft robotic muscle was based on the fingers used in P16227, with no significant alterations.

Soft Robotic Muscle Version 2

public/Photo Gallery/New Actuator.PNG

The second version of the soft robotic muscle is altered to account for both attachment to the robot and ease in pneumatic assembly.

McKibben Muscle

public/Photo Gallery/McKibben.PNG

The McKibben muscle design utilizes coupling hex nuts and a quick-disconnect for tubing.

Interchangeable connections

Bone model with connection points, displays the interchangeability of the leg design where all muscles and bones can be isolated and replaced. Each leg will have 4 muscles (A,B,C, and D) that will be attached in these general locations. The push to connect will allow the "Bone" to be detached from the air flow system with ease.
public/Photo Gallery/Velcro.PNG
Hook and loop strips with adhesive backs can be attached to the bones or the body, and used to secure the muscles. This way, muscles are easily removed from the system.
public/Photo Gallery/MTC.PNG
The MettleAir quick disconnect provides convenient detachment from the air system if necessary.
public/Photo Gallery/Barb.PNG
Barbed tube fittings could allow an easy disconnect from the air flow system for the soft robotic muscles.

Weight Estimation

Created by Naveena Shanmugam
Components Weight (oz) Weight per Qty (lb) Qty Weight Total (lb)
Battery 80 5 1 5
PCB + Arduino + Xbox receiver 24 1.5 1 1.5
Air Tank 9.6 0.6 1 0.6
Air Compressor 48 3 1 3
Solenoids 2.2 0.1375 19 2.6125
Tubing 3.2 0.2 1 0.2
Bone 0.2 0.1 8 0.8
McKibben Muscles 1.2 0.6 8 4.8
Silicon Muscles 1.4 0.0875 8 0.7
Regulator 6 0.375 2 0.75
Manifold for Solenoid 4.6 0.2875 1 0.2875
Total Weight 20.25

Center of Mass

A rough, preliminary center of mass analysis. Analysis is based on the assumption that onboard component mass is centered on body. Importantly, we can see that center of mass will be higher on robot than desired, as currently designed.


Pneumatic System This is a system level diagram of the air flow that includes the basic components. All solenoid valves will be normally closed valves. Air flows from the compressor to the tank through a check valve and 3-way solenoid valve 1. During operation solenoid valve A will be in the actuated position. This ensures that if there is any electrical failure to the system the air safely vents to atmosphere instead of exploding and components. Solenoid valve 2 will be in the un-actuated position when the system is on in order to allow the air tank to fill before walking. Once optimal pressure is reached solenoid valve 2 will remain open for the duration of robot use unless there is a critical decrease in pressure. Once solenoid valve 2 opens air will pass through regulator one in order to reduce air pressure and flow. Solenoid valve 3 will open allowing the reduced airflow into the bone structure. The valve will then close once the desired PSI has been reached(TBD). Solenoid valves 4 through 11 will remain closed until the bone structure is inflated. Air will flow from the tank through regulator 2 and produce a reduced airflow and pressure. Solenoid valves 4 through 11 will then open allowing the muscle system to inflate to base pressure(TBD). Solenoid valves 4 through 11 will then open and close based on the user input from the controller and allow the robot to walk. Detailed preliminary schematic of the air flow system. Appropriate pressure fittings, various connectors, and additional valves will be added once they are determined.


Initial Solenoid Test Schematic

The purpose of this schematic is to test one instance of interfacing with the solenoids used to control air flow. Using a supply of 24 V to power the solenoid and a 5 V supply to act as the Arduino, the 5 V signal shall be toggled on the gate of the NMOS to control the solenoid. This shall be accomplished via breadboard and once verified manually, the Arduino shall be used to control one solenoid before moving to 19 instances (expected number of needed solenoids).

Ardunio Solenoid Example

12 V Power Supply Testing

In order to operate the robot, the air compressor must be powered and preferably on-board. Progressing forward with a 6S LiPo battery, the necessary DC-DC converter must step down the nominal 22.2 V battery to 12 V at 11 A. Through research and consulting a local technical salesman, the part capable of achieving these specifications is LTC3807. In order to test this part, its associated demo board provided by Linear Technologies (DC2221A) shall be utilized.

Using the aforementioned battery or an equivalent voltage, the board shall be tested to verify its response to large loads (power resistor first, then air compressor). Once verified, the design shall progress forward by interfacing with other subsystems and further schematic development. It is expected that this demo board will provide adequate voltage (12 V) and current (<15 A).

LTC3807 Product Page

Prior to any physical testing, simulations were performed using LTSpice. The circuit tested may be seen below.

The values for all of the components were calculated via LTpowerCAD. This resulted in the waveform as seen below with Vout (green) and Iout (blue).

As verified above, the circuit is capable of providing 12 V at 15 A. Physical testing will further verify this design.

Low Voltage Cutoff Simulations

Due to the sensitivity of lithium polymer batteries, they can not be discharged below 2.8 V per cell absolute minimum. In order to prevent this, a low voltage cutoff circuit was developed to disable any load from further depleting the battery. Using the voltage comparator LTC1540, the circuit will turn off at 3.2 V per cell for safe measure. The circuit tested may be seen below.

In the simulations, the relay was replaced with a equivalent resistive load to provide the required current to activate the relay. By sweeping the voltage V1, the effect of a depleting battery can be observed below.

At a voltage of 19.2 V, the equivalent voltage per cell of a 6 cell battery is 3.2 V, which verifies the concept.

Bill of Material (BOM)

Bill of Materials

Test Plans

Bone Load Bearing Test Plan

To be performed by Avery Becker.

Objective/Data to Acquire: Determine the relationship between the pressure of the “Inflatable Bones” and load required for cause the bone to buckle.


Procedure: Normal Stress

Per table one below, the bone prototype will be inflated to the desired psi. After inflation, the bone will be place onto the compression rig. Weights will be added in 5 lb increments until the prototype buckles and fails. The weight before buckling will be recording and any other observations (buckle location, air leakage, etc) will be noted.

Table 1

Pressure Load Comments

Procedure: Shear Stress

Per table two below, the bone prototype will be inflated to the desired psi. After inflation, the bone will be place onto the shear rig. Weights will be added in 1 lb increments on the free end of the rig until the prototype buckles and fails. The weight before buckling will be recording and any other observations (buckle location, air leakage, etc) will be noted.

Table 2.

Pressure Load Comments

Leg Load Bearing Test Plan

Leg Testing Plan

Muscle Force Output Test Plan

Muscle Testing Plan

Electrical Hardware/Software Testplan(s)

Solenoid Driver Test
  1. To be performed by: T. Brudz
  2. Equipment to be used: power supply, breadboard, necessary components, oscilloscope, digital multimeter.
  3. Data to be acquired: waveforms of applying voltage to gate of NMOS device, resulting current of applying voltage to gate of NMOS device

12 V Power Supply Test

  1. To be performed by: T. Brudz
  2. Equipment to be used: DC2221A demo circuit, oscilloscope, digital multimeter, power supply.
  3. Data to be acquired: waveform of voltage and associated ripple under light (200 mA) and heavy load (12 A)

Software Test Plan

Objective: To make have a safe first interface with the hardware


  1. To be tested by: Naveena Shanmugam
  2. Equipment to be used: Arduino, Xbox controller, LEDs, project hardware
  3. Test all hardware connected to digital I/O pins of Arduino using dc voltages to ensure correct implementation
  4. Connect LEDs to used digital I/O pins of Arduino and test the various software commands one by one
  5. Connect Arduino to project hardware setup and ensure that the various software commands interface with the hardware properly and work as expected

Design and Flowcharts

This section should continue to be updated from your systems level design documentation.

Predetermined pattern of Movement

Risk Assessment

Plans for next phase

Trevor's Three Week Plan

Avery's Three Week Plan

Amanda's Three Week Plan

Nicole's Three Week Plan

Cameron's Three Week Plan

Naveena's Three Week Plan

Home | Planning & Execution | Imagine RIT

Problem Definition | Systems Design | Preliminary Detailed Design | Detailed Design

Build & Test Prep | Subsystem Build & Test | Integrated System Build & Test | Integrated System Build & Test with Customer Demo | Customer Handoff & Final Project Documentation