An Autonomous, Optically-Powered, Direct-to-Digital Sun-Angle Recorder for Honey Bee Flight Tracking
An Autonomous, Optically-Powered, Direct-to-Digital Sun-Angle Recorder for Honey Bee Flight Tracking
复制标题
DOI：
10.1109/tcsii.2021.3067033
复制
发表时间：
2021-05-01
影响因子：
4.4
通讯作者：
Molnar, Alyosha C.
中科院分区：
文献类型：
--
作者：
Palmer, Daniel M.;Molnar, Alyosha C.
介绍了一种用于昆虫飞行跟踪的自主传感器，该传感器无需使用传统的ADC即可捕获和存储太阳入射角。该芯片使用角度敏感像素对作为差分1位ADC，这些传感器阵列提供入射角度的多位编码。通过使用新的像素内比较器，立即在像素内进行数字化，从而实现超低功率、直接到数字的入射角捕获。这些测量值在整个飞行过程中存储在芯片存储器中，并可以在飞行结束时通过磁后向散射上传到基站。为了实现完全自主，该系统由芯片上的光伏供电，在自然阳光下提供0.40-0.50V的开路电压，其强度在整个飞行过程中可以改变。在额定电源电压为0.45V的情况下，系统的模拟功耗为630nW，并给出了适用于光伏电源条件的电路拓扑。该芯片被设想为一个安装在蜜蜂身上的飞行记录器，用于跟踪蜜蜂的飞行轨迹，通过携带芯片沿226米的室外轨迹飞行，并在野外的基站从芯片下载记录的飞行数据来模拟蜜蜂的飞行。在相同轨迹的18次试验中，模拟标记蜜蜂访问已知喂食点并返回蜂箱的小样本，重建的平均最终位置距离真实最终位置9.6米，误差不到总轨迹长度的5%。
An autonomous sensor for insect flight tracking that captures and stores solar angle-of-incidence without the use of a conventional ADC is presented. The chip uses pairs of Angle-Sensitive Pixels as differential 1-bit ADCs, and an array of these sensors provides a many-bit encoding of angle-of-incidence. Digitization occurs immediately within the pixel by use of a novel in-pixel comparator, thus enabling ultra-low-power, direct-to-digital capture of angle-of-incidence. These measurements are stored in an on-chip memory throughout the flight and can be uploaded to a base station via magnetic backscatter at the end of the flight. To permit full autonomy, the system is powered by on-chip photovoltaics, providing open-circuit voltages in the range of 0.40 - 0.50 V in natural sunlight, the intensity of which can vary throughout a flight. The system consumes a simulated 630nW at nominal supply voltage of 0.45 V, and circuit topologies amenable to photovoltaic power conditions are presented. The chip is envisioned as a flight recorder to be mounted on honey bees to track their trajectories, and bee flights were emulated by carrying the chip along a 226 m outdoor trajectory and downloading the recorded flight data from the chip at a base station in the field. Across 18 trials of the same trajectory, which emulates a small sample of tagged bees visiting known feeding sites and returning to their hive, the average reconstructed final position is 9.6 m away from the true final position, an error less than 5% of the total trajectory length.