TinySense: A Lighter Weight and More Power-efficient Avionics System for Flying Insect-scale Robots

TL;DR

The paper tackles sensor autonomy for sub-gram flying insect robots by introducing TinySense, an ultra-lightweight avionics suite that enables hover without external sensing. It replaces a laser rangefinder with a pressure sensor and moves optic-flow processing onboard a global-shutter camera, fused via a Kalman Filter using a minimal state $\boldsymbol{q}=[\theta, v_x, z]^T$ and measurements $\Omega_m = \omega_m - \dfrac{v_x}{z_d}$. Key contributions include reducing the sensor payload to $78.4\mathrm{ mg}$ and power to about $15\mathrm{ mW}$ (with optic-flow compute ~5 mW), achieving RMSEs of $\text{pitch}=1.573^{\circ}$, $\text{velocity}=0.186$ m/s, and $\text{altitude}=0.136$ m in flight relative to motion capture. The work demonstrates practical sensor autonomy within tight energy budgets and lays the groundwork for insect-scale autonomous navigation and swarm-scale deployment, while outlining directions to handle fast maneuvers and vibration in real-world platforms.

Abstract

In this paper, we introduce advances in the sensor suite of an autonomous flying insect robot (FIR) weighing less than a gram. FIRs, because of their small weight and size, offer unparalleled advantages in terms of material cost and scalability. However, their size introduces considerable control challenges, notably high-speed dynamics, restricted power, and limited payload capacity. While there have been advancements in developing lightweight sensors, often drawing inspiration from biological systems, no sub-gram aircraft has been able to attain sustained hover without relying on feedback from external sensing such as a motion capture system. The lightest vehicle capable of sustained hovering -- the first level of ``sensor autonomy'' -- is the much larger 28 g Crazyflie. Previous work reported a reduction in size of that vehicle's avionics suite to 187 mg and 21 mW. Here, we report a further reduction in mass and power to only 78.4 mg and 15 mW. We replaced the laser rangefinder with a lighter and more efficient pressure sensor, and built a smaller optic flow sensor around a global-shutter imaging chip. A Kalman Filter (KF) fuses these measurements to estimate the state variables that are needed to control hover: pitch angle, translational velocity, and altitude. Our system achieved performance comparable to that of the Crazyflie's estimator while in flight, with root mean squared errors of 1.573 deg, 0.186 m/s, and 0.136 m, respectively, relative to motion capture.

Figures (5)

• Figure 1: The TinySense sensor suite shown next to a U. Washington Robofly and a standard pencil for scale. The width of the sensor suite board is approximately 12 mm and its mass is 78.4 mg.
• Figure 2: Forces and state variables of a small hovering Flying Insect Robot (FIR).
• Figure 3: Sensor suite (PAG7920LT optical sensor, ICM42688-P IMU, BMP388 pressure sensor) next to US 1 cent coin.
• Figure 4: Sensor measurements from the onboard gyroscope, optic flow sensor, and pressure sensor of TinySense are compared with those from the Crazyflie for three different flight experiments (a), (b), and (c); trajectories are described in the main text and corresponding state estimates in Fig. \ref{['state estimation']}. In the altitude plots, altitude is measured by the Crazyflie's laser rangefinder, while the TinySense uses a pressure sensor; mocap data is also included for comparison.
• Figure 5: Comparison between state estimates from TinySense's Kalman Filter, Crazyflie, and mocap. We conducted three flight experiments (a), (b), and (c), with the Kalman filter estimates starting at time zero. Pressure sensor measurements were ignored during the first 1.8 s because pressure changes due to the ground effect at takeoff confound readings.