BiPneu: Design and Control of a Bipolar-Pressure Pneumatic System for Soft Robots

Yu Mei, Xinyu Zhou, Vedant Naik, Alan Gao, Xiaobo Tan · 2026

Read BiPneu: Design and Control of a Bipolar-Pressure Pneumatic System for Soft Robots by Yu Mei, Xinyu Zhou, Vedant Naik, Alan Gao, Xiaobo Tan (2026) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2605-12804.txt.

Introduction

Mei and four colleagues present BiPneu, a bipolar-pressure system for soft parts. Two-way pressure regulation is key to soft parts, giving wide motion. Yet regulation across both sides is hard, due to uneven moves, valve kinks and flow jolts. Their DM-SMC controller uses hysteresis choice from a mixed air-electric model. It beats model and PID controls, cutting error by 11.9 percent in step tests and 35.6 percent in sine tests.

Catalog record

CREATORYu Mei, Xinyu Zhou, Vedant Naik, Alan Gao, Xiaobo Tan
YEAR2026
COLLECTIONPAPERS
FORMATTXT
FILE SIZE57141
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2605.12804

Browse the full library · About the Tower of Babel

Full text

The opening of the work, from the library's hosted text:

BiPneu: Design and Control of a Bipolar-Pressure Pneumatic System for Soft Robots Yu Mei, Xinyu Zhou, Vedant Naik, Alan Gao, Xiaobo Tan Abstract Positive–negative pressure regulation is critical to soft robotic actuators, enabling large motion ranges and versatile actuation modes. However, achieving high-performance regulation across both pressure polarities remains challenging due to asymmetric inflation–deflation dynamics, valve nonlinearities, and switching-induced flow disturbances. This paper presents BiPneu , a scalable and cost-efficient multi-channel bipolar-pressure pneumatic system for soft robots that enables wide-range, accurate, and responsive pressure regulation while providing seamless compatibility with high-level software ecosystems. A dual-mode sliding-mode controller (DM-SMC) with hysteresis-supervised mode selection is proposed based on a hybrid electro-pneumatic model. Extensive simulation and experiments demonstrate the superior performance of DM-SMC in tracking step and sinusoidal pressure references compared with both advanced model predictive controllers and well-tuned PID controllers. Experimental results show average absolute errors of 1.44 kPa in multi-step tests and 4.23 kPa in sinusoidal tracking, corresponding to reductions of 11.9% and 35.6% relative to PID control, along with improved control effort, valve switching rate, and transient response. Robustness of DM-SMC is further verified on a bellow actuator with pressure-dependent volume. Finally, BiPneu’s capability is demonstrated via two soft robotic examples, quick ball-maneuvering with a soft parallel manipulator and real-time finite element method (FEM)-based teleoperation of a soft bellows actuator. I Introduction Soft robots have attracted increasing attention for their inherent compliance and adaptability, enabling safe exploration in unstructured environments [ 7 ] and robust interaction with humans [ 6 , 9 ] . Among various actuation technologies for soft robots, pneumatic actuation is particularly appealing because it offers high power-to-weight ratio and fast response with low-cost hardware. More recently, a growing class of bipolar-pressure (i.e., positive–negative) pneumatic actuators has emerged [ 3 , 6 ] . These actuators exploit both positive and negative pressures to realize bidirectional motion, larger workspace, and versatile actuation modes. Despite these benefits, achieving rapid, precise, and robust pressure control across bipolar pressure regimes is challenging due to asymmetric inflation/deflation dynamics and strong valve nonlinearities. The soft robotics community has developed an extensive set of pneumatic actuation systems, which can be broadly categorized into three classes [ 2 ] : volume control, continuous pressure control using proportional valves, and switching pressure control using solenoid on/off valves. Volume control uses syringes or piston pumps to vary the chamber volume and induce deformation [ 3 ] . However, displacement sources often provide limited airflow and are typically bulky. For continuous pressure control, proportional valves enable accurate regulation by accepting continuous analog voltages, eliminating the need for high-frequency pulse-width modulation (PWM) control [ 12 , 1 , 16 ] . These valves can be easily driven by microprocessor units, such as Raspberry Pi. Such platforms natively support standard network protocols (e.g., TCP/IP) and robotic middleware frameworks (e.g., ROS 2), enabling seamless integration with high-level control and robotic software ecosystems. However, proportional valves are generally much larger and more expensive than solenoid on/off valves (often 2–5 \times in size and 3–20 \times in cost) [ 2 , 16 ] , limiting compact and scalable implementations for robots requiring many independently controlled channels. In contrast, switching-valve architectures regulate pressure using PWM-driven solenoid on/off valves, enabling relatively compact and cost-effective multi-channel implementation. Consequently, a variety of on/off–valve-based pneumatic systems have been developed for soft robots [ 8 , 5 ] . However, existing platforms primarily target positive-pressure regulation, with only a few supporting bipolar pressure operation [ 13 , 2 , 17 , 14 ] . Among these bipolar pressure systems with on/off valves, some realize regulation using per-channel mini pumps [ 13 , 17 , 14 ] , which increases size and complexity and often constrains the achievable pressure range and transient response. Another type of bipolar pressure systems with on/off valves rely on low-level microcontroller units (MCUs) for PWM generation, which can hinder integration with high-level software stacks. Overall, existing platforms rarely achieve bipolar-pressure regulation within a cost-effective on/off–valve–based architecture, while supporting integration with high-level robotic software ecosystems. In addition to hardware design, achieving high-performance control in on/off–valve-based bipolar pressure systems is difficult. While existing bipolar pressure platforms typically adopt PID for pressure regulation [ 2 , 17 ] , inflation–deflation switching induces transient flow disturbances and asymmetric dynamics due to air compressibility and valve nonlinearities, often exciting nonlinear oscillatory pressure responses that render fixed-gain PID tuning difficult and insufficient [ 8 ] . To address this challenge, prior work has explored enhanced strategies such as fuzzy piecewise PID [ 13 ] , dual-loop PID tuned via evolutionary algorithms [ 8 ] , and model-based reinforcement learning [ 12 ] . While model-based control offers an appealing alternative, most efforts on characterization and modeling of pneumatic platforms are limited to individual components [ 18 ] . System-level modeling and closed-loop regulation of pneumatic systems remain largely unexplored. TABLE I : Comparison of our work with the state of the art in bipolar pressure pneumatic systems. Description Characteristics of n channels Reference [ 17 ] Reference [ 14 ] Reference [ 2 ] Reference [ 13 ] Reference [ 1 ] Our work Number of channels ( n ) 6 10 8 \leq 8 \leq 12 16 Air source n mini pumps 2n mini pumps 1 air pump 1 vacuum pump n mini pumps 1 air pump 1 vacuum pump 1 air pump 1 vacuum pump Valves 3n on–off valves n on–off valves 2n on–off valves 3n on–off valves n proportional valves 2n on–off valves Control system n Arduino 1 PCB 1 MCU 1 PCB 1 Raspberry Pi 1 Raspberry Pi Regulation range -59 – 112 kPa -50 – 80 kPa -50 – 150 kPa -60 – 100 kPa -100 – 100 kPa -80 – 200 kPa (up to 300 kPa) Control accuracy (w.r.t. regulation range) <3 kPa (1.7%) <0.2 kPa (0.2%) <1 kPa (0.5%) <1.5 kPa (1.0%) <1 kPa (0.5%) <1.5 kPa (0.5%) Response time① >1.5 s >0.7 s <0.2 s 0.6 s 0.06 – 0.27 s 0.13 s Software ecosystems compatibility② ✗ ✗ ✗ ✗ ✓ ✓ Size 30 \times 24 \times 11 cm 32 \times 23 \times 12 cm 24 \times 16 \times 11 cm 10 \times 10 \times 2 \times n cm 15 \times 13 \times 10 cm 25 \times 25 \times 19 cm Cost per channel③ Medium Low Low Low High Low ① Measured using a fixed 20 mL air volume with a pressure step of 0\rightarrow 100 kPa; ② Supports TCP/IP networking and ROS 2–based robotic middleware; ③ Cost categories: High > $500, Medium $100–$500, and Low < $100. In this work, we introduce BiPneu , a scalable and cost-efficient multi-channel bi polar-pressure pneu matic system that enables accurate and responsive regulation of both positive and negative pressures over a wide operating range, while providing compatibility with high-level robotic software ecosystems. A comparison between BiPneu and representative state-of-the-art bipolar-pressure pneumatic systems is summarized in Table I . In our preliminary work [ 10 ] , we developed a hybrid electro-pneumatic model and a mixed-integer nonlinear MPC (MI-NMPC) framework that embeds the discrete inflation and deflation mode into the predictive-control formulation to co-optimize mode scheduling and PWM inputs. In this paper we substantially extend that

Read the complete text