Self‐Powered Bio‐Inspired Spider‐Net‐Coding Interface Using Single‐Electrode Triboelectric Nanogenerator

Abstract Human–machine interfaces are essential components between various human and machine interactions such as entertainment, robotics control, smart home, virtual/augmented reality, etc. Recently, various triboelectric‐based interfaces have been developed toward flexible wearable and battery‐less applications. However, most of them exhibit complicated structures and a large number of electrodes for multidirectional control. Herein, a bio‐inspired spider‐net‐coding (BISNC) interface with great flexibility, scalability, and single‐electrode output is proposed, through connecting information‐coding electrodes into a single triboelectric electrode. Two types of coding designs are investigated, i.e., information coding by large/small electrode width (L/S coding) and information coding with/without electrode at a predefined position (0/1 coding). The BISNC interface shows high scalability with a single electrode for detection and/or control of multiple directions, by detecting different output signal patterns. In addition, it also has excellent reliability and robustness in actual usage scenarios, since recognition of signal patterns is in regardless of absolute amplitude and thereby not affected by sliding speed/force, humidity, etc. Based on the spider‐net‐coding concept, single‐electrode interfaces for multidirectional 3D control, security code systems, and flexible wearable electronics are successfully developed, indicating the great potentials of this technology in diversified applications such as human–machine interaction, virtual/augmented reality, security, robotics, Internet of Things, etc.

The corresponding generated peaks when finger sliding forward and backward across the eight electrode patterns.
The photograph of a fabricated eight-direction BISNC interface of L/S coding design is shown in Figure S2a, with large electrode width of 8 mm, small electrode width of 4 mm, and electrode spacing of 6 mm. According to the design principle, electrode with larger width should produce higher output peak due to the larger contact area during sliding. In the measurement, the sliding motion is performed by finger with normal speed to slide forward (inside toward outside) and backward (outside toward inside) across the electrode patterns.
The reason for backward sliding is because it is difficult to decode all the directions only based on forward sliding generated peaks with small spacing between the electrodes.
Due to the relatively large size of human finger of ~15 mm and the small electrode spacing of 6 mm, there are situations where finger are covering two electrodes (e.g., finger sliding out of the first electrode and sliding in the second electrode) at the same time. This causes the overlapping of negative component of the first peak and positive component of the second peak, leading to a reduction in amplitude for the second peak, as shown in Figure   S2b-i. Due to the overlapping of output peaks, the output signal pattern may not accurately follow the electrode pattern in an ideal case (i.e., larger output signal peak from electrode with larger width and smaller output signal peak from electrode with smaller width).
Therefore, a forward/backward sliding and a detection strategy are proposed for the interpretation of the signal peaks corresponding to the electrode pattern, as illustrated in Table   1. The last two columns of the table indicate the comparison results of the current peak with the former peak (i.e., the second peak with the first peak, and the third peak with the second peak). The comparison results are roughly categorized into three classes, i.e., larger ("L"), equivalent ("E") and smaller ("S"). With the overlapping effect, the large/small (former electrode is large and current electrode is small) comparison result is always "S", same as the electrode pattern. Similarly, results of large/large and small/small can be "E" or "S", while results of small/large comparison can be "L" or "E". For the patterns with only two strip 5 electrodes (direction 1 and 8), only one comparison is required for forward and backward sliding. Based on the comparison results as indicated in the table, they can be easily differentiated. For the other six directions with three strip electrode in the pattern, whenever an "L" comparison result appears in the forward and backward sliding, it means that the current electrode must have a larger electrode width than the former one. That is, the current electrode width is large, and the former electrode width is small. If both the forward sliding and backward sliding have "L", the electrode pattern can be easily interpreted, such as direction 3 and 6. Then the other directions can be interpreted according to the comparison results in the table S1. Table S1. Signal interpretation table (comparison result of the current peak amplitude with the former one: "L" -larger, "E" -equivalent, "S"smaller).

Direction
Electrode patterns  In the case of 3 peaks, the ideal signal (100) with constant speed is shown in Figure S3a, with black peak representing the references peak, solid red peak representing the real coding peak and dash red peak representing the virtual coding peak generated from the corresponding electrodes. In the schematic, d is the distance between two adjacent electrodes, v1 is the average sliding speed from the beginning reference peak to the signal peak, v2 is the average sliding speed from the signal peak to the ending reference peak, T1 and T2 are the time duration. The dash lines in the figure denote the range of the generated signal peak for correct recognition. For 100, T1 should be less than 3/8 of the entire time duration between the two reference peaks. After the calculation, the relationship of v1 and v2 can be achieved, i.e., v2 < 9/5 • v1. For 010, T1 should locate in between 3/8 and 5/8 of the entire time duration, thus 3/5 • v1 < v2 < 5/3 • v1. Similarly, for 001, T1 should be larger than 5/8 of the entire time duration, thus v2 > 5/9 • v1. In summary, the variation of sliding speed should satisfy the condition of 3/5 • v1 < v2 < 5/3 • v1.
In the case of 4 peaks, the ideal signal (110) with constant speed is shown in Figure S3b.
For 110, to achieve correct recognition, first T3 should be larger than both T1 and T2, and then it should be also larger 3/8 of the entire time duration. Therefore, by considering all the scenarios and conditions, the variation of sliding speed should within the range from 60% to 166.7%, in order to achieve correct recognition. That is, the 0/1-coding control interface has a variation tolerance in sliding speed of at least ±40%.  Video S1. Demonstration of 3D drone control.

S5. Trend of time intervals for rotation and up
Video S2. Operation of the flexible BISNC interface.
Video S3. Operation of the stretchable BISNC interface.