Search for: All records

Aerial insects exhibit agile maneuvers such as sharp braking, saccades, and body flips under disturbances; in contrast, insect-scale aerial robots are limited to tracking smooth trajectories with small acceleration. To achieve similar flight capabilities, insect-scale robots require a robust and computationally efficient controller. Here, through designing a deep-learned robust tube model predictive controller, we showcase exceptional flight agility in a 750-milligram flapping-wing robot. Our neural network controller can track aggressive trajectories and run at a high rate on a compute-constrained system. The robot demonstrates saccades with a lateral speed and acceleration of 197 centimeters per second and 11.7 meters per square second, respectively, representing improvements of 447 and 255% over prior results. The robot also performs saccades under 160–centimeters per second wind disturbance and completes 10 consecutive somersaults in 11 seconds. These results represent a milestone in achieving insect-scale flight agility and inspire future investigations on sensory and compute autonomy. 

Insect-scale robots face two major locomotive challenges: constrained energetics and large obstacles that far exceed their size. Terrestrial locomotion is efficient yet mostly limited to flat surfaces. In contrast, flight is versatile for overcoming obstacles but requires high power to stay aloft. Here, we present a hopping design that combines a subgram flapping-wing robot with a telescopic leg. Our robot can hop continuously while controlling jump height and frequency in the range of 1.5 to 20 centimeters and 2 to 8.4 hertz. The robot can follow positional set points, overcome tall obstacles, and traverse challenging surfaces. It can also hop on a dynamically rotating plane, recover from strong collisions, and perform somersaults. Compared to flight, this design reduces power consumption by 64 percent and increases payload by 10 times. Although the robot relies on offboard power and control, the substantial payload and efficiency improvement open opportunities for future study on autonomous locomotion. 

Aerial insects are exceptionally agile and precise owing to their small size and fast neuromotor control. They perform impressive acrobatic maneuvers when evading predators, recovering from wind gust, or landing on moving objects. Flapping-wing propulsion is advantageous for flight agility because it can generate large changes in instantaneous forces and torques. During flapping-wing flight, wings, hinges, and tendons of pterygote insects endure large deformation and high stress hundreds of times each second, highlighting the outstanding flexibility and fatigue resistance of biological structures and materials. In comparison, engineered materials and microscale structures in subgram micro–aerial vehicles (MAVs) exhibit substantially shorter lifespans. Consequently, most subgram MAVs are limited to hovering for less than 10 seconds or following simple trajectories at slow speeds. Here, we developed a 750-milligram flapping-wing MAV that demonstrated substantially improved lifespan, speed, accuracy, and agility. With transmission and hinge designs that reduced off-axis torsional stress and deformation, the robot achieved a 1000-second hovering flight, two orders of magnitude longer than existing subgram MAVs. This robot also performed complex flight trajectories with under 1-centimeter root mean square error and more than 30 centimeters per second average speed. With a lift-to-weight ratio of 2.2 and a maximum ascending speed of 100 centimeters per second, this robot demonstrated double body flips at a rotational rate exceeding that of the fastest aerial insects and larger MAVs. These results highlight insect-like flight endurance, precision, and agility in an at-scale MAV, opening opportunities for future research on sensing and power autonomy. 

Insects can navigate in cluttered spaces and perform challenging functions such as pollination and collective object transport. By exploiting scaling laws and bioinspired designs, insect‐scale micro‐aerial‐vehicles (MAVs) have demonstrated impressive flight capabilities such as in‐flight collision resilience and acrobatic maneuvers. However, existing subgram MAVs are difficult to design, construct, and repair. Coupled with challenges in robot sensing and control, existing subgram MAVs have not achieved insect‐like swarm flight, which limits potential studies of swarm behaviors and future applications such as collective sensing. Herein, a new design and fabrication method is developed to substantially improve the fabrication scalability of subgram MAVs. Based on a small set of design parameters, an automated algorithm generates the laser cut files of microrobotic components. To reduce fabrication and assembly time, stereolithographic 3D printing is used for making static components such as the airframe and connectors. The modular design enables straightforward assembly and repair, which reduces the overall fabrication time by over 2 times. Owing to the ease of fabrication and good reliability, two subgram MAVs demonstrate controlled hovering flight and coordinated lifting of an object. This result lays the foundation for future robotic studies of collective insect flight. 

Soft-actuated insect-scale micro aerial vehicles (IMAVs) pose unique challenges for designing robust and computationally efficient controllers. At the millimeter scale, fast robot dynamics (∼ms), together with system delay, model uncertainty, and external disturbances significantly affect flight performances. Here, we design a deep reinforcement learning (RL) controller that addresses system delay and uncertainties. To initialize this neural network (NN) controller, we propose a modified behavior cloning (BC) approach with state-action re-matching to account for delay and domain-randomized expert demonstration to tackle uncertainty. Then we apply proximal policy optimization (PPO) to fine-tune the policy during RL, enhancing performance and smoothing commands. In simulations, our modified BC substantially increases the mean reward compared to baseline BC; and RL with PPO improves flight quality and reduces command fluctuations. We deploy this controller on two different insect-scale aerial robots that weigh 720 mg and 850 mg, respectively. The robots demonstrate multiple successful zero-shot hovering flights, with the longest lasting 50 seconds and root-mean-square errors of 1.34 cm in lateral direction and 0.05 cm in altitude, marking the first end-to-end deep RL-based flight on soft-driven IMAVs. 

The increasing popularity of video streaming and conferencing services have altered the nature of Internet traffic. In this paper, we take a first step toward quantifying the impact of this changing nature of traffic on the Quality of Experience (QoE) of popular video streaming and conferencing applications. We first analyze the traffic characteristics of these applications and of backbone links, and show how simple multipath routing may adversely impact application QoE. To mitigate this problem, we propose a new routing path selection approach, inspired by the TCP timeout computation algorithm, that uses both the average and variation of path load. Preliminary results show that this approach improves application QoE by on average 14% and packet latency by 11% for video streaming and conferencing applications, respectively.