Sensing systems typically use separate photoelectric sensors for non-contact perception and pressure sensors for contact perception. Because these sensors are spatially separated and often acquire signals at different times, their data must be aligned and fused afterward, introducing potential latency, positional mismatch, and signal crosstalk.

In a study published in Nature Sensors, a team led by Prof. CHEN Ming and Prof. YANG Chunlei from the Shenzhen Institute of Advanced Technology of the Chinese Academy of Sciences, along with Prof. WEI Lei's team from Nanyang Technological University, Singapore, developed a bio-inspired flexible bimodal sensor that enables photoelectric and pressure sensing at the same location and at the same time.

Inspired by the tube feet of sea urchins which integrate photoreceptive and mechanosensitive functions within a single biological structure, the researchers developed a vertically integrated flexible sensor which consists of two vertically integrated functional layers.

The upper SnSexSy/PTAA heterojunction provides broadband photoelectric sensing, while the lower covalently interlocked PP/FCNT conductive network enables sensitive and highly linear pressure detection. The two layers share an electrode and employ independent dual-channel signal readout to minimize interference.

The sensor achieved a pressure sensitivity of 245.41 kPa^-1 with a linearity of R² > 0.99 over 0-40 kPa, and a photoresponsivity of 85.31 A/W. Cross-channel interference between photoelectric and pressure signals was below 1%.

A 5×5 sensor array using the sensor recognized 10 types of objects with 96% accuracy, higher than using optical sensing alone and using pressure sensing alone. A spherical robot equipped with 36 sensors achieved a 94% success rate in mapless autonomous navigation in a simulated fire environment. A system using the sensor simultaneously tracked soil moisture and light intensity, achieving over 99% recognition accuracy across three soil types.

The "homologous perception" strategy enables optical and mechanical information to originate from the same physical location, providing inherently aligned signals without relying on post-hoc fusion. This study provides a new paradigm for multimodal flexible sensing, with potential applications in embodied intelligence, human–robot interaction, and precision agriculture.

In the future, researchers plan to pursue miniaturization and large-scale array integration, with a focus on dexterous manipulation and tactile sensing.