昆虫スケールの触覚知覚とナビゲーションのための生物模倣ロボ物理アンテナの設計
Design of a bioinspired robophysical antenna for insect-scale tactile perception and navigation
ゴキブリの触角を模倣し、軽量・低消費電力の多関節柔軟センサCITRASを開発。触覚情報から距離や隙間幅の推定、表面テクスチャの識別を実現した。
詳しい要約
1. どんなもの?
2. 先行研究と比べてどこがすごい?
3. 技術・手法の肝は?
4. どうやって有効だと検証した?
5. 議論はある?
6. 次に読むべき論文は?
※ AIが要旨から生成した要約です。正確性は原文をご確認ください。
著者: Parker McDonnell, Lingsheng Meng, Hari Krishna Hariprasad, Alexander Hedrick, Eduardo Miscles, Samuel Gilinsky, Jean-Michel Mongeau, Kaushik Jayaram
分類: cs.RO
原文アブストラクト
The American cockroach (Periplaneta americana) uses its soft antennae to guide decision making by extracting rich tactile information from tens of thousands of distributed mechanosensors. Although tactile sensors enable robust, autonomous perception and navigation in natural systems, replicating these capabilities in insect-scale robots remains challenging due to stringent size, weight, and power constraints that limit existing sensor technologies. To overcome these limitations, we introduce CITRAS (Cockroach Inspired Tactile Robotic Antenna Sensor), a bioinspired, multi-segmented, compliant laminate sensor with embedded capacitive angle sensors. CITRAS is compact (73.7x15.6x2.1 mm), lightweight (491 mg), and low-power (32 mW), enabling seamless integration with miniature robotic platforms. The segmented compliant structure passively bends in response to environmental stimuli, achieving accurate hinge angle measurements with maximum errors of just 0.79 degree (quasistatic bending) and 3.58 degree (dynamic bending). Experimental evaluations demonstrate CITRAS' multifunctional tactile perception capabilities: predicting base-to-tip distances with 7.75 % error, estimating environmental gap widths with 6.73 % error, and distinguishing surface textures through differential sensor response. The future integration of this bioinspired tactile antenna in insect-scale robots addresses critical sensing gaps, promising enhanced autonomous exploration, obstacle avoidance, and environmental mapping in complex, confined environments.
