Fully autonomous drones are a new emerging field that has enabled many applications such as gas source leakage localization, wild-fire detection, smart agriculture, and search and rescue missions in unknown limited communication and GPS denied environments. Artificial intelligence and deep Neural Networks (NN) have enabled applications such as visual perception and navigation which can be deployed to make drones smarter and more efficient. However, deploying such techniques on tiny drones is extremely challenging due to the limited computational resources and power envelope of edge devices. To achieve this goal, this paper proposes an efficient end-to-end optimization method for deploying deep NN models for visionbased autonomous drone navigation applications, such as obstacle avoidance and steering task. This paper formulates two different methods for implementing the NN inference phase onto tiny drones and analyzing the implementation results for each case: 1) a Cloud-IoT implementation and 2) Onboard Processing. Several models are trained with state-of-the-art scalable NN architectures and the most efficient cases in terms of computation complexity and accuracy are selected for implementation on a cloud server and several edge devices. By designing hardware-friendly NN models and optimal configuration of the implementation platforms, we were able to reach up to 97% accuracy, speed up the computation 2.3x, have 22x less complexity, and 53% energy reduction. Also, we achieve up to 25 fps on the GAP8 processor, which is enough for real-time drone navigation requirements, even when the model is running on a small IoT device.

In this paper, we propose an energy-efficient architecture which is designed to receive both images and text inputs as a step towards designing reinforcement learning agents that can understand human language and act in real-world environments. We evaluate our proposed method on three different software environments and a low power drone named Crazyflie to navigate towards specified goals and avoid obstacles successfully. To find the most efficient language-guided RL model, we implemented the model with various configurations of image input sizes and text instruction sizes on the Crazyflie drone GAP8 which consists of 8 RISC-V cores. The task completion success rate and onboard power consumption, latency, and memory usage of GAP8 are measured and compared with Jetson TX2 ARM CPU and Raspberry Pi 4. The results show that by decreasing 20% of input image size we achieve up to 78% energy improvement while achieving an 82% task completion success rate.

Mixed-Criticality Systems (MCSs) include tasks with multiple levels of criticality and different modes of operation. These systems bring benefits such as energy and resource saving while ensuring safe operation. However, management of available resources in order to achieve high utilization, low power consumption, and required reliability level is challenging in MCSs. In many cases, there is a trade-off between these goals. For instance, although using fault-tolerance techniques, such as replication, leads to improving the timing reliability, it increases power consumption and can threaten life-time reliability. In this work, we introduce an approach named Life-time Peak Power management in Mixed-Criticality systems (LPP-MC) to guarantee reliability, along with peak power reduction. This approach maps the tasks using a novel metric called Reliability-Power Metric (RPM). The LPP-MC approach uses this metric to balance the power consumption of different processor cores and to improve the life-time of a chip. Moreover, to guarantee the timing reliability of MCSs, a fault-tolerance technique, called task re-execution, is utilized in this approach. We evaluate the proposed approach by a real avionics task set, and various synthetic task sets. The experimental results show that the proposed approach mitigates the aging rate and reduces peak power by up to 20.6% and 17.6%, respectively, compared to state-of-the-art.

No abstract

Safety, low-cost, small size, and Artificial Intelligence (AI) capabilities of drones have led to the proliferation of autonomous tiny Unmanned Aerial Vehicles (UAVs) in many applications which are dangerous, unknown, or time-consuming for humans. Deep Neural Networks (DNNs) have enabled autonomous navigation while using captured data by drone sensors as input to the model. Due to the extreme complexity of DNNs, cloud-based approaches have been highly addressed in which a drone is connected to the cloud and sends the data to the cloud, and takes the result. On the other hand, emerging tiny machine learning models and edge computing brings significant improvement in energy efficiency and latency with respect to cloud-based approaches. However, there is a trade-off in these two implementations for model accuracy, latency, and energy efficiency. For instance, applying tiny machine learning models leads to lower latency but it sacrifices model accuracy in comparison to cloud-based computing. To address these challenges, we consider multiple models and introduce a new approach named MLAE2 which applies Metareasoning approach for Latency-Aware Energy-Efficient autonomous drones. Metareasoning monitors parameters such as latency and energy consumption for different algorithms and chooses the appropriate algorithm due to the environmental situation changes. To Evaluate our approach we extract the power consumption and latency for both cloudbased computing and edge computing while deploying multiple models on a tiny drone named Crazyflie. The experimental results show that MLAE2 successfully meets the latency constraint while maximizing model accuracy and improving energy efficiency.