SYNFORMAX

TECHNOLOGY Materials science, mechatronics & embodied intelligence

From flexible actuators to full‑stack control —
systems that move the way the world actually is,
not the way rigid machines wish it were.

Explore ↓
The Core Premise

Every technology category below exists to answer
the same question: what happens when a machine
can actually feel what it's doing?


Not sensing through a camera from a distance.
Not force-limited after the fact.
Understanding through contact.

That is the design problem we are solving.

Core Capabilities

What We
Build

01

Soft Grippers

Adaptive elastomer fingers with integrated tactile sensing. Tool-free module swaps and quick-change couplers mean zero retooling when the SKU changes. The fingers read what they hold and adjust grip force in real time.

Pneumatic · Cable-Driven · Tactile Feedback
02

Artificial Muscles

HASEL and DEA actuation for lifelike motion profiles. Lightweight, scalable, and compliant by default — no gears, no bearings, no rigid linkages. The mechanism is the material.

HASEL · DEA · Electroactive Polymer
03

Embodied AI

Perception-in-the-loop control on ROS 2. Learn motion primitives by demonstration; deploy as safe, repeatable skills. The system improves with every cycle it runs.

ROS 2 · Motion Learning · Sensor Fusion
04

Standards & Safety

Designed with ANSI/RIA and ISO collaborative robot standards from day one. Instrumented testing and risk assessment built into the development cycle — not retrofitted at the end.

ANSI/RIA · ISO 10218 · TS 15066
05

Integration

Drop-in end-effectors for popular collaborative robots. Python and C++ APIs. Web dashboard for pilot operations. Designed to fit into existing infrastructure, not replace it.

Python · C++ · REST API · Web Dashboard
06

Grant-Ready R&D

Research architecture aligned with NSF, DARPA, and DOE funding tracks. Modular platforms designed for academic and national lab collaboration from the first prototype.

NSF · DARPA · DOE · Open Architecture
Technical Detail

How Each
System Works

Soft Grippers

The Compliant
End-Effector

Traditional rigid grippers succeed at one task: gripping objects that are already the right shape. Our elastomer finger modules use pneumatic and cable-driven actuation to conform to whatever they encounter. Integrated tactile sensors at each fingertip close the loop — the system knows not just that it's holding something, but how hard, where, and whether the grip is stable.

Quick-change couplers allow finger modules to be swapped in under 60 seconds without tools. Different finger geometries ship for different task profiles — pinch, wrap, lateral — each hot-swappable on the same base.

Actuation: Pneumatic + Cable · Sensing: Capacitive Tactile · Swap Time: <60s
Artificial Muscles

The Actuator
as Material

HASEL (Hydraulically Amplified Self-healing Electrostatic) actuators use a liquid dielectric encased in a flexible pouch. Apply voltage — the pouch deforms, producing contraction or extension. No motors. No gearboxes. No rigid linkages.

DEA (Dielectric Elastomer Actuator) modules offer a complementary profile: higher strain, lower force, ideal for large-displacement applications like artificial limbs or soft locomotion. Both actuator families are self-damping — they absorb impact energy rather than transmitting it.

HASEL: ~25% strain · DEA: ~40% strain · Both: self-healing, silent, shock-resilient
Embodied AI

The Control
Architecture

The control stack runs on ROS 2 with real-time perception pipelines that feed learned motion policies. Rather than hand-coding trajectories, operators demonstrate a task and the system extracts a generalizable motion primitive — a skill that transfers across variations in object position, size, and orientation.

Deployed skills run as safe, interruptible behaviors. The system monitors force, velocity, and contact state continuously. If any threshold is exceeded, the behavior pauses gracefully rather than faulting hard. Skills improve with every successful execution logged.

Stack: ROS 2 · Learning: Imitation + RL · Safety: Continuous force/velocity monitoring
Integration

Deploying into
What You Have

End-effector modules ship with standardized mounting flanges for UR, Fanuc, KUKA, and ABB collaborative robots. The Python SDK exposes full actuator and sensor state with a clean async API. A web dashboard provides real-time monitoring, skill library management, and pilot logging without requiring a local install.

For custom deployments, the C++ library provides deterministic real-time control at 1kHz update rates. All interfaces are documented and versioned — no black boxes, no proprietary lock-in.

Mounting: ISO 9409-1 · SDK: Python + C++ · Dashboard: Web · Latency: 1kHz
Compliance & Interoperability

Built to
Standard

Safety Standards
ANSI/RIA R15.06Industrial robots — safety requirements for integration and use in collaborative applications.
ISO 10218-1 / -2Robots and robotic devices — safety requirements for industrial robots, parts 1 and 2.
ISO/TS 15066Collaborative robots — power and force limiting, speed and separation monitoring, hand-guiding.
IEC 61508Functional safety of electrical/electronic/programmable electronic safety-related systems.
Integration & Interfaces
ISO 9409-1 FlangeStandardized robot tool mounting — compatible with all major cobot arms out of the box.
ROS 2 (Humble / Iron)Full ROS 2 node and action server interfaces. Drop into any existing ROS 2 workspace.
Python & C++ SDKAsync Python client for rapid prototyping. Deterministic C++ library for real-time control at 1kHz.
REST / WebSocket APILanguage-agnostic control and monitoring. Web dashboard ships with zero local dependencies.
Ready to Go Deeper

See the technology
in your application.

We run pilots with logistics operators, research institutions, and healthcare partners. If you have a handling problem that rigid grippers keep failing at — let's talk.

Start a Conversation → Grant One‑Pager → Pitch One‑Pager
No lock-in. No proprietary hardware dependency. Just results.