The invisible connection: RIKEN’s stretchable film and the next challenge for wearable electronics

The component most likely to disappear from a picture of wearable technology may be the one that determines whether it works: the connection. A sensor can be exceptionally thin, yet still need a dependable route to its power supply and readout circuit. Every turn of the wrist asks that route to accommodate movement while keeping the electrical signal intact.

RIKEN reported a new approach on September 19: an approximately 300-nanometre anisotropic conductive film, made with silver nanowires and an elastic polymer, joined at room temperature without applied heat or pressure using a volatile liquid. Its Japanese announcement reports conductivity beyond 500% tensile strain, water-vapour transmission of 1,200 grams per square metre per day, stability after 1,000 stretching cycles, and working light-emitting and wrist-mounted connections. The team includes researcher Sunghoon Lee and team director Takao Someya. The work appeared in Science Advances.^[1]

The overlooked interface

The practical significance lies in a question larger than any single material: what happens between the devices? Consider a future skin patch that combines a compliant sensing surface with a separate readout module. Its designer must provide a connection without making that junction the part that restricts movement. The relevant object is the assembled system, including the places where one material ends and another begins.

Anisotropic conduction means electrical behaviour depends on direction. For a connector, the useful distinction is between carrying current across the intended contact and accidentally allowing it to travel into a neighbouring circuit. More conductivity everywhere would not necessarily make a better connector. The pattern of insulation matters alongside the conducting pathways.

This is an engineering interpretation of the research, rather than a claim that a finished consumer device is ready. It changes where to look for progress: inside the joints, as well as on the sensing surface.

A history of making electronics less obtrusive

In July 2013, the Japan Science and Technology Agency and the University of Tokyo announced a two-micrometre organic transistor circuit and demonstrated a touch-sensor application. Crucially, the stated thickness excluded external power and display units.^[2] That qualification remains useful today. The dimensions of an active sheet are not automatically the dimensions of everything needed to operate it.

Later that month, another announcement described an ultrathin organic light-emitting device that continued working when bent.^[3] Taken together, those developments illustrate different parts of the same ambition: sensing and displaying information on surfaces that do not behave like a rigid circuit board. Connecting such parts creates its own set of demands.

The historical lesson is not that thinness alone solves wearability. It is that each successful component exposes the next integration problem. A sensor that follows the body still needs power, an output and a way to survive assembly. Progress in one area makes shortcomings elsewhere more consequential.

Why breathability became part of the story

A 2017 University of Tokyo research announcement described conductive nanomesh electronics, including a week-long skin-patch test and electromyogram recordings.^[4] RIKEN’s contemporary account explained why this mattered: even thin polymer supports could trap moisture beneath them.^[5] The question had broadened from whether electronics could conform to skin to what covering skin meant over time.

That earlier work supplies context, not a transferable safety certificate. Different materials and structures need their own evaluation. A favourable result for one mesh cannot establish how every later connector will behave on every wearer.

For designers, the distinction is substantial. A patch that feels unobtrusive at the moment of application may be used in conditions quite unlike that first demonstration. Comfort belongs in the design brief, but it must also be defined: during what activity, over what area, and for how long?

Reading performance numbers without turning them into promises

The useful way to read a material specification is to attach a question to it. Water-vapour transmission asks how much vapour passes through an area over time under the measurement conditions. It is not, by itself, a measure of how much sweat a complete product can manage or how many days someone should wear it.

A complete patch could also cover the skin with other components, protective layers and attachment materials. Assessing the connector separately helps identify its contribution; assessing the assembled patch asks a different question. Neither measurement substitutes for the other.

Likewise, a large single deformation and repeated smaller movements examine different aspects of durability. A cycle count without the accompanying deformation, speed and environmental conditions cannot be converted into a service life. Nor should the largest reported strain automatically be assumed to apply to every repeated-cycle test.

Room-temperature joining also describes a particular assembly step. It does not mean the entire manufacturing process consumes no energy, that no solvent is involved, or that an eventual electronic system produces no heat. Keeping those boundaries explicit makes the achievement easier to assess.

Everyday conditions are a separate design challenge

A separate paper published in npj Flexible Electronics in February 2026 examined nanomesh electrodes combining polyvinyl alcohol and waterborne polyurethane. The authors reported at least four hours of stable electrical and mechanical behaviour on the palm, a site exposed to perspiration and motion.^[6] This was a different electrode system, but it illustrates why breathability, water resistance and deformation must be considered together.

For a hypothetical exercise-monitoring patch, the critical question would be whether useful signals remain available during the activity of interest. A device might deliver accurate readings while the wearer rests and lose contact at the moment of exertion. Its evaluation should therefore consider the gaps in the record as well as the quality of the readings retained. This is an application-level inference, not a reported clinical outcome of the new film.

For longer use, developers would also need to define how the device is attached, removed and replaced. Those ordinary actions can be central to whether a technically promising system becomes practical.

What the idea could mean for soft robots

Imagine a soft gripper equipped with touch sensors. Information must travel from a deforming surface to control electronics. The connection has to fit the mechanical design as well as the electrical one. A compliant connector could expand the designer’s choices about where to place those interfaces.

But a robotic application is not simply a skin patch with a different label. Its loads, abrasion, cleaning procedures and consequences of failure may differ. Before selecting a connector, a developer would need to specify the movement, current and operating environment it must withstand. An attractive laboratory result identifies an option to investigate; it does not establish suitability for every machine.

The path from one working sample to repeatable assembly

For manufacturers, the next questions concern consistency. Can contacts be aligned repeatedly? At what stage can an incomplete connection be detected? Does storage change the assembly process? Can a failed component be replaced without discarding the rest of the device? These are practical evaluation questions, not reported defects in the research.

For health applications, reliable interconnection would still be only one part of the evidence. Signal quality, interpretation and the usefulness of the resulting information would have to be established for the intended task. A connector cannot, on its own, validate a diagnosis.

The advance is best understood within that longer sequence: make electronics thin enough to follow a surface, consider what covering that surface does, and develop ways to connect the parts without surrendering their useful properties. The joint may be nearly invisible in the finished device. It deserves a prominent place in judging whether the device can work.