Soft Robotics: Molding, Arduino & Closed-Loop Control

Learning Goal: Design, mold, and test a pneumatic soft robotic gripper using silicone elastomer, implementing a closed-loop pressure control system with an Arduino and solenoid valves to safely grasp delicate objects.

Prerequisites: Basic understanding of algebra and general physics concepts. No prior programming or molding experience is required.
Estimated Total Study Time: 25 Hours

Module 1: Introduction to Soft Robotics & Pneumatic Actuation

This module introduces the interdisciplinary field of soft robotics. You will explore how compliant, bio-inspired materials differ from traditional rigid systems, learn the fundamental physics of fluidic elastomer actuators (specifically Pneunets), and examine real-world applications in handling delicate objects.

💡 Supplement: Actuator Bending Physics & Independent Study

While these high-level overviews demonstrate soft robotic capabilities, they do not illustrate the exact internal chamber expansion physics of Pneunet bending actuators.

How a Pneunet (Pneumatic Network) Actuator Works: A Pneunet actuator consists of two primary layers:

The Expandable Chambers (Top Layer): Molded from a soft, stretchable silicone elastomer (low durometer, e.g., Ecoflex).
The Strain-Limiting Layer (Bottom Layer): Reinforced with an unstretchable material (such as paper, fabric, or a high-durometer, stiff silicone like Dragon Skin).

When air is pumped into the internal chambers, the pressure forces the thin walls of the top chambers to expand outward. Because the bottom layer is constrained by the strain-limiting material, it cannot stretch longitudinally. This asymmetry in material strain forces the entire actuator to bend toward the strain-limiting side.

   UNPRESSURIZED                       PRESSURIZED (Bending)
 _________________                      _____      _____
[ | | | | | | | | ] Chambers           (     )____(     )  <-- Expandable
=================== Strain-Limiter      \==============/   <-- Restrained

Independent Search Recommendation: To visualize this expansion in real-time, search YouTube for: "Pneunet bending actuator working principle animation" or "Harvard soft robotics toolkit pneumatic gripper simulation".

Knowledge Checkpoint

Explain the difference between "structural compliance" and "rigid joint actuation."
Identify the two layers required to make a soft actuator bend instead of inflating like a balloon.
Define what a "strain-limiting layer" is and name two materials commonly used to create it.
Describe why soft actuators are inherently safer for handling fragile structures compared to metal parallel grippers.

Module 2: Arduino Programming and Electronics Foundations

To build an automated, closed-loop pressure system, you must first master the electronic control framework. This module covers the basics of Arduino microcontrollers, safe breadboard prototyping, and using NPN transistors and N-channel MOSFETs to switch higher-power devices (like 12V air pumps and solenoids) from 5V digital pins.

Knowledge Checkpoint

Draw a diagram showing how breadboard terminal strips are electrically connected relative to the power rails.
Write a simple Arduino sketch that reads an analog input pin and prints the raw 10-bit integer value (0–1023) to the Serial Monitor.
Identify the three main pins of a transistor/MOSFET (Gate/Base, Drain/Collector, Source/Emitter) and explain their function in switching a 12V load.
Explain why a flyback diode is necessary when switching inductive loads like solenoid valves and motors.

Module 3: Silicone Elastomer Molding and Fabrication

Note: Following the revised curriculum sequence, this module has been moved ahead of pneumatic circuit design. Silicone elastomers require between 4 and 24 hours to fully cure at room temperature. By fabricating and casting your soft fingers now, you can leave them to cure while you study circuit design and sensor integration in Module 4.

Knowledge Checkpoint

Describe the mixing ratio of EcoFlex (e.g., EcoFlex 00-30 or 00-50) and explain why precise, slow mixing is critical.
Explain how a vacuum chamber removes microscopic air bubbles from viscous silicone mixtures.
Outline the process for bonding the strain-limiting bottom layer to the chambered top layer without clogging the internal air pathways.
Identify two signs of under-cured silicone and name two factors (such as cold temperatures or sulfur contamination) that can inhibit curing.

Module 4: Pneumatic Circuit Design and Solenoid Control

In this module, you will design the physical pneumatic system that controls the airflow. You will build a hardware circuit that interfaces an Arduino with a 12V mini air pump, electronic solenoid valves (to direct inflation, holding, and venting), and an analog electronic pressure sensor to read system pressure in real-time.

🛠️ Pneumatic Circuit Schematic

To control your gripper, assemble your hardware according to this schematic:

        +12V DC Power Supply
             |
             +---------------+---------------+
             |               |               |
         [12V Pump]    [12V Solenoid]  [Pressure Sensor VCC]
             |               |               |
             + (Flyback      + (Flyback      |
             |  Diodes)      |  Diodes)      |
             |               |               |
          (Drain)         (Drain)            |
         [MOSFET 1]      [MOSFET 2]          |
          (Source)        (Source)           |
             |               |               |
 GND --------+---------------+---------------+ (GND Sensor)
  |          |               |               |
  |          |               |               |

Knowledge Checkpoint

Explain the difference between a 2-way/2-position (normally closed) solenoid valve and a 3-way/2-position valve in a pneumatic circuit.
Write a line of C++ code to scale a raw 10-bit analog read (0-1023) from a 0.5V-4.5V pressure sensor (0 to 100 kPa) into an actual pressure float variable in kPa.
Draw a diagram showing how to connect a flyback diode across a inductive load (pump/valve coil) to prevent transient voltage spikes from damaging your MOSFET.
Explain why a solid-state MOSFET can handle high-frequency switching, whereas a mechanical relay cannot.

Module 5: Closed-Loop Pressure Control & Grasp Testing

This final module ties your physical silicone actuators, pneumatic hardware, and electronics together. You will implement a feedback loop (using Proportional-Integral-Derivative logic, or a simplified threshold-based hysteresis loop) to inflate your soft gripper to a precise target pressure, holding it steady to safely grasp fragile items without crushing them.

🎓 Supplement: Gap-Filling Arduino PID Implementation

As noted in the course review, few online tutorials provide complete, ready-to-use code for managing pressure systems using simple solenoid valves. To bridge this gap, write your code using the structured framework below. This example implements a threshold-based hysteresis controller to inflation-limit your system, protecting fragile objects like eggs.

Arduino Soft Robotics Closed-Loop Pressure Control Framework
Safely inflation-limits and holds a target pressure within a soft finger. */

const int SENSOR_PIN = A0; // Pressure sensor analog output const int PUMP_PIN = 5; // MOSFET Gate driving the 12V Air Pump const int VALVE_PIN = 6; // MOSFET Gate driving the 12V Vent Solenoid Valve

// Target and safety limits (values in kPa) float targetPressure = 15.0; // Target pressure for secure grasp float safetyPressureLimit = 25.0; // Hard cutoff to prevent actuator rupture float pressureHysteresis = 1.5; // Acceptable deadband around target pressure

void setup() { pinMode(PUMP_PIN, OUTPUT); pinMode(VALVE_PIN, OUTPUT); Serial.begin(9600);

// Safe default: system vented, pump off digitalWrite(PUMP_PIN, LOW); digitalWrite(VALVE_PIN, HIGH); // Assuming normally-closed valve: HIGH opens/vents }

float readPressureKPa() { int rawADC = analogRead(SENSOR_PIN); // Example scaling formula: 0.5V - 4.5V sensor mapped to 0 - 100 kPa float voltage = rawADC * (5.0 / 1023.0); float pressure = (voltage - 0.5) * (100.0 / 4.0); if (pressure < 0) pressure = 0; return pressure; }

void loop() { float currentPressure = readPressureKPa();

// --- Safety Override --- if (currentPressure >= safetyPressureLimit) { digitalWrite(PUMP_PIN, LOW); // Force pump off digitalWrite(VALVE_PIN, HIGH); // Open vent valve immediately Serial.println("!!! OVERPRESSURE SAFETY LIMIT EXCEEDED. VENTING !!!"); while(true); // Lock execution for manual safety reset }

// --- Hysteresis Loop Control --- if (currentPressure < (targetPressure - pressureHysteresis)) { // Under-pressurized: Inflate digitalWrite(VALVE_PIN, LOW); // Close venting valve digitalWrite(PUMP_PIN, HIGH); // Run pump Serial.print("Status: INFLATING... "); } else if (currentPressure > (targetPressure + pressureHysteresis)) { // Over-pressurized but under safety limit: Vent slightly digitalWrite(PUMP_PIN, LOW); // Turn off pump digitalWrite(VALVE_PIN, HIGH); // Open vent valve slightly Serial.print("Status: DEFLATING... "); } else { // Within acceptable target range: Hold pressure digitalWrite(PUMP_PIN, LOW); // Turn off pump digitalWrite(VALVE_PIN, LOW); // Close valve to trap air inside finger Serial.print("Status: HOLDING TARGET... "); }

Serial.print("Pressure (kPa): "); Serial.println(currentPressure); delay(100); // 10Hz control loop cycle }

🥚 Experimental Testing Protocol: Safe Egg Grasping

Once your physical setup and code are ready, use this step-by-step testing protocol to evaluate your system's performance safely:

Dry-Run Calibration: Connect your cured fingers to the pneumatic circuit without placing any objects in the grip. Run your code with a very low target pressure (e.g., 5 kPa). Verify that your pump runs, the fingers bend inward, the sensor reads the rising pressure, and the pump cuts off when the target is reached.
Leak Testing: Set the target pressure to 12 kPa. Once reached, observe the system status. If the pump cycles on and off repeatedly while trying to hold pressure, look for tiny air leaks along your tubing connectors or silicone seams using a soapy water solution (which will bubble at leak sites).
Rigid Target Trial: Position a rigid plastic bottle inside the gripper. Run the system to verify that the fingers conform around the container and hold pressure securely.
Fragile Target Test (The Egg Trial):
- Use an raw chicken egg as your test object. Mount it securely in a vertical stand.
- Start with a conservative target pressure of 8 kPa. Run the system.
- Observe if the fingers conform to the egg's surface. Try gently lifting the gripper base. If the egg slips, increase the target pressure in 1.5 kPa steps until the grasp is secure.
- Record the minimum holding pressure ( $P_{hold}$ ) that successfully lifts the egg without damage.
- Verify that your software safety limit is set to no more than $2 \times P_{hold}$ to prevent accidental crushing.

Knowledge Checkpoint

Identify the three terms of a PID controller ( $K_p$ , $K_i$ , $K_d$ ) and describe how each influences how quickly and stably your system reaches its target pressure.
Explain what "integral windup" is and how it can cause a pressure system to overshoot and damage a fragile object.
Describe the purpose of a "hysteresis band" (deadband) in a simple threshold-based valve controller.
List two methods for physically sealing air connections between stiff vinyl tubing and flexible silicone fingers to prevent system leaks.

Course Map

This flowchart maps the recommended path through the modules. Notice how Module 3 (Silicone Molding) is scheduled before Module 4 (Pneumatics) to allow curing time, helping you make the most of your lab hours.

Key People Index

Harvard Biodesign Lab Researchers (The Soft Robotics Toolkit Team): An open-source collective that helped standardize soft robotic mold designs and pneumatic control systems, making soft robotics fabrication accessible to students and hobbyists worldwide.
The RBO Lab Team (TU Berlin): Creators of the highly resilient PneuFlex soft actuator designs. They pioneered techniques for incorporating structural fabric constraints directly into molded silicones.

Final Course Self-Assessment

Complete this checklist to verify that you have mastered all the skills taught in this curriculum:

Explain the physics behind bending actuators (specifically the relationship between expandable chambers and strain-limiting layers).
Choose an appropriate silicone elastomer (such as EcoFlex 00-30 or 00-50) based on its durometer, pot life, and cure time.
Mix, degas in a vacuum chamber, cast, and cure a leak-free multi-chambered silicone actuator finger.
Wire a safe breadboard circuit that connects an Arduino to high-power 12V pneumatic devices using MOSFETs and flyback diodes.
Interface an analog pressure transducer with an Arduino, read its output voltages, and mathematically scale them to physical units of pressure.
Implement an Arduino control loop (such as a hysteresis or PID loop) that dynamically monitors pressure sensor readings and controls air valves.
Configure code-level safety overrides that dump system pressure immediately if readings exceed safety limits.
Safely grasp and lift a delicate, raw egg using your closed-loop pneumatic soft robotic gripper system.

Explore Further

Related Engineering Roadmaps

View All→

engineering

Soft Robotics: Molding, Arduino & Closed-Loop Control

Module 1: Introduction to Soft Robotics & Pneumatic Actuation

Recommended Videos

💡 Supplement: Actuator Bending Physics & Independent Study

Knowledge Checkpoint

Module 2: Arduino Programming and Electronics Foundations

Recommended Videos

Knowledge Checkpoint

Module 3: Silicone Elastomer Molding and Fabrication

Recommended Videos

Knowledge Checkpoint

Module 4: Pneumatic Circuit Design and Solenoid Control

Recommended Videos

🛠️ Pneumatic Circuit Schematic

Knowledge Checkpoint

Module 5: Closed-Loop Pressure Control & Grasp Testing

Recommended Videos

🎓 Supplement: Gap-Filling Arduino PID Implementation

🥚 Experimental Testing Protocol: Safe Egg Grasping

Knowledge Checkpoint

Course Map

Key People Index

Final Course Self-Assessment

Related Engineering Roadmaps

ECG Monitor Design: Op-Amps, Filters & Noise

Building EMG Prosthetics: Arduino & 3D Print

Building a CNC Plotter: GRBL & G-Code Setup

Planetary Gearboxes: CAD, Backlash & Torque