Video calls solved sight and sound and left touch entirely alone. The attempts to fill that gap have mostly used vibration, because vibration motors are cheap, small and easy to drive — and vibration does not feel like being touched. It feels like a phone.
Hugging Suit: Pneumatically-Actuated System Design for Remote Haptic Experiences, posted 6 October 2026 by Russian (Ruo-Xuan) Wu, Luke Hespanhol, Marius Hoggenmueller, Hannes Waldschütz and Eva Hornecker, uses pressure instead.
The system
Three components:
- An Air Actuator Matrix (AAM) — a programmable pneumatic array that delivers pressure
- A portable high-pressure pneumatic unit — the air supply
- A fabric-based pressure sensor layer — enabling two-way haptic feedback
Described as precise, wearable haptic feedback while remaining lightweight and mobile, and maker-friendly in design, with applications in long-distance relationships, emotional support and therapeutic contexts.
Why pneumatics rather than vibration
The choice is the engineering argument, and it comes down to what skin actually senses.
Different mechanoreceptors respond to different things. Vibration at 100–300Hz is detected primarily by Pacinian corpuscles, which are tuned to rapid transients — they are the receptors that register a buzz, a texture scraping past, a tool hitting something. Sustained pressure over an area is sensed by Merkel cells and Ruffini endings, which respond to static indentation and skin stretch.
A hug is sustained, distributed pressure. It is the second channel almost entirely, which is why a vibrating garment cannot approximate one no matter how it is programmed. You are stimulating the wrong receptors.
Pneumatics give you that channel directly. An inflating bladder produces exactly what pressure receptors are built to detect: area contact, gradual onset, sustained force, slow release.
The trade is in the hardware. Pneumatics are bulky, need a compressor or a reservoir, and are slow — inflation and deflation take hundreds of milliseconds. “Portable high-pressure pneumatic unit” is the paper doing the difficult part, and lightweight and mobile is the claim to interrogate, because every previous pneumatic haptic suit has been tethered to a bench.
And the fabric sensor layer is what makes it a channel rather than a broadcast. One-directional haptics is a notification. Two-way means the person hugging feels themselves hug — the squeeze registers, and so the gesture has the feedback that makes it feel performed rather than sent.
The finding worth taking away
Low-lit, private spaces, especially when users could also see their remote partner via video, improved emotional engagement and comfort during mediated haptic experiences.
This is not a hardware result and it is the most actionable thing in the paper.
Context did as much work as the actuators. The same suit, delivering the same pressure, worked better in a dim private room with the partner visible than it did otherwise. Which says that the thing being engineered is not a sensation — it is an experience that the sensation is one component of.
Three reasons that makes sense:
Low light reduces competing input. Touch is a low-bandwidth channel relative to vision. In a bright, visually busy room, a subtle pressure cue has to compete for attention; in the dark it is most of what is happening.
Privacy removes self-consciousness. Being hugged by a machine in front of other people is a performance. Alone, it can be received.
And video supplies the referent. Pressure on its own is pressure. Pressure while you can see the person who sent it is a hug, because the sight provides the attribution that makes the sensation meaningful. The haptics are not communicating who — the video is doing that, and the touch is carrying intensity and timing.
For anyone building this kind of work
Design the room, not just the device. This is the transferable lesson and it applies far beyond haptics. An installation’s lighting, privacy and social configuration will frequently affect the experience more than another iteration of the hardware. For a piece involving touch, intimacy or vulnerability, the room is most of it.
And “maker-friendly” is worth taking seriously. Soft pneumatic actuators are genuinely within reach — silicone casting, TPU bladders, a small 12V pump, solenoid valves and a microcontroller gets you a working single-channel system for under a hundred pounds. The soft-robotics literature is open and well documented, and the hard part is not the pneumatics but the valve manifold that lets one pump address many actuators independently, which is what an “air actuator matrix” is.
Worth reading alongside our force-sensing primer — the fabric pressure sensor layer here is the input side of the same physics.