When testing conditions differ from those represented in training data, so-called out-of-distribution (OOD) inputs can mar the reliability of learned components in the modern robot autonomy stack. Therefore, coping with OOD data is an important challenge on the path towards trustworthy learning-enabled open-world autonomy. In this paper, we aim to demystify the topic of OOD data and its associated challenges in the context of data-driven robotic systems, drawing connections to emerging paradigms in the ML community that study the effect of OOD data on learned models in isolation. We argue that as roboticists, we should reason about the overall \textit{system-level} competence of a robot as it operates in OOD conditions. We highlight key research questions around this system-level view of OOD problems to guide future research toward safe and reliable learning-enabled autonomy.

Drone-to-drone detection using visual feed has crucial applications, such as detecting drone collisions, detecting drone attacks, or coordinating flight with other drones. However, existing methods are computationally costly, follow non-end-to-end optimization, and have complex multi-stage pipelines, making them less suitable for real-time deployment on edge devices. In this work, we propose a simple yet effective framework, \textit{TransVisDrone}, that provides an end-to-end solution with higher computational efficiency. We utilize CSPDarkNet-53 network to learn object-related spatial features and VideoSwin model to improve drone detection in challenging scenarios by learning spatio-temporal dependencies of drone motion. Our method achieves state-of-the-art performance on three challenging real-world datasets (Average Precision@0.5IOU): NPS 0.95, FLDrones 0.75, and AOT 0.80, and a higher throughput than previous methods. We also demonstrate its deployment capability on edge devices and its usefulness in detecting drone-collision (encounter). Project: \url{https://tusharsangam.github.io/TransVisDrone-project-page/}.

There is a great demand for vision-based robotics solutions that can operate using Global Positioning Systems (GPS), but are also robust against GPS signal loss and gyroscope failure. This paper investigates the estimation and tracking control in application to a Vertical Take-Off and Landing (VTOL) Unmanned Aerial Vehicle (UAV) in six degrees of freedom (6 DoF). A full state observer for the estimation of VTOL-UAV motion parameters (attitude, angular velocity, position, and linear velocity) is proposed on the Lie Group of $\mathbb{SE}_{2}\left(3\right)\times\mathbb{R}^{3}$ $=\mathbb{SO}\left(3\right)\times\mathbb{R}^{9}$ with almost globally exponentially stable closed loop error signals. Thereafter, a full state observer-based controller for the VTOL-UAV motion parameters is proposed on the Lie Group with a guaranteed almost global exponential stability. The proposed approach produces good results without the need for angular and linear velocity measurements (without a gyroscope and GPS signals) utilizing only a set of known landmarks obtained by a vision-aided unit (monocular or stereo camera). The equivalent quaternion representation on $\mathbb{S}^{3}\times\mathbb{R}^{9}$ is provided in the Appendix. The observer-based controller is presented in a continuous form while its discrete version is tested using a VTOL-UAV simulation that incorporates large initial error and uncertain measurements. The proposed observer is additionally tested experimentally on a real-world UAV flight dataset. Keywords: Unmanned aerial vehicle, nonlinear filter algorithm, autonomous navigation, tracking control, feature measurement, observer-based controller, localization, exponential stability, asymptotic stability, inertial measurement unit (IMU), Global Positioning Systems (GPS), vision aided inertial navigation system.

Wing, Alphabet's aviation subsidiary, is partnering with Walmart to kick off drone deliveries from the retail chain in the Dallas-Fort Worth (DFW) metro area. The flights will begin taking off "in the coming weeks" from a Walmart Supercenter in Frisco, TX, and the companies plan to expand to a second DFW location before the end of the year. The companies say the coverage area from both stores will cover 60,000 homes. The service will be available to homes within about six miles of the supported stores. Residents in those areas can order things like quick meals, groceries, essentials and over-the-counter medicines. The drones can fly up to 65 mph, and Wing says you'll get your items in under 30 minutes.

Turkish drone attacks in northern Iraq have killed seven members of the Kurdistan Workers' Party (PKK), authorities said, as the country's foreign minister met the president and prime minister of Iraq's semi-autonomous Kurdistan Regional Government (KRG). "A Turkish army drone struck a PKK vehicle, killing an official and two fighters", the KRG's counterterrorism services said on Thursday. The attack took place in Sidakan district, north of the regional capital Erbil. Later, the counter-terrorism services said that another drone strike in Sidakan had killed four PKK members, including two medical personnel. The PKK has fought a rebellion against Turkey since 1984, and has bases inside KRG territory.

A Ukrainian strike destroyed a missile complex in Russian-occupied Crimea on Wednesday, August 23, Ukraine's military intelligence agency said. Russia and Ukraine launched drone strikes against each other Wednesday morning, each looking to score a major win in a fight that continues to drag on with little progress or end in sight. Ukrainian intelligence claimed to have destroyed a Russian S-400 surface-to-air missile defense system in Crimea, while Russia struck grain facilities in Odesa overnight Tuesday. The S-400 system shows another instance of Ukraine's plan to strike at Russian assets, even behind the front line. Ukraine's intelligence agency GUR claimed on its Telegram channel that Russia has a "limited number" of sophisticated systems left and that this loss strikes a "painful blow" to their forces.

Russian air defence systems have taken down three unmanned aerial vehicles (UAV) that tried to attack Moscow, the latest raid on Russia's capital by combat drones that authorities have accused Ukraine of launching. Russia's defence ministry said one drone was jammed electronically and crashed into a building in central Moscow early on Wednesday morning, and two more were shot down by air defence systems outside the capital. Moscow's Mayor Sergei Sobyanin said on the Telegram messaging app that one downed drone had hit a building that was under construction in central Moscow, and another was shot down in a district to the west of the city. The second UAV hit a building under construction in the City," Sobyanin said on Telegram. Russia's defence ministry said that the third drone was shot down in the Khimki district of Moscow.

General Oleksandr Tarnavskyi, the deputy commander of Ukrainian forces in the south, said Ukraine's troops had gained a footing in the southeastern village of Robotyne and were organising the evacuation of civilians. Oleksandr Prokudin, the governor of Ukraine's Kherson region, said an elderly woman was killed and a 55-year-old man injured in Russian air attacks. A drone raid was reported in Moscow, forcing a temporary halt to air traffic at Vnukovo, Sheremetyevo and Domodedovo airports. City Mayor Sergei Sobyanin said Russian air defence systems shot down the two drones west of the capital and blamed Ukraine. Russia's Air Force said it scrambled two jets against two drones flying near the Crimean Peninsula, which it annexed in 2014.

A reliable communication network is essential for multiple UAVs operating within obstacle-cluttered environments, where limited communication due to obstructions often occurs. A common solution is to deploy intermediate UAVs to relay information via a multi-hop network, which introduces two challenges: (i) how to design the structure of multi-hop networks; and (ii) how to maintain connectivity during collaborative motion. To this end, this work first proposes an efficient constrained search method based on the minimum-edge RRT$^\star$ algorithm, to find a spanning-tree topology that requires a less number of UAVs for the deployment task. To achieve this deployment, a distributed model predictive control strategy is proposed for the online motion coordination. It explicitly incorporates not only the inter-UAV and UAV-obstacle distance constraints, but also the line-of-sight (LOS) connectivity constraint. These constraints are well-known to be nonlinear and often tackled by various approximations. In contrast, this work provides a theoretical guarantee that all agent trajectories are ensured to be collision-free with a team-wise LOS connectivity at all time. Numerous simulations are performed in 3D valley-like environments, while hardware experiments validate its dynamic adaptation when the deployment position changes online.

The real-time dynamic environment perception has become vital for autonomous robots in crowded spaces. Although the popular voxel-based mapping methods can efficiently represent 3D obstacles with arbitrarily complex shapes, they can hardly distinguish between static and dynamic obstacles, leading to the limited performance of obstacle avoidance. While plenty of sophisticated learning-based dynamic obstacle detection algorithms exist in autonomous driving, the quadcopter's limited computation resources cannot achieve real-time performance using those approaches. To address these issues, we propose a real-time dynamic obstacle tracking and mapping system for quadcopter obstacle avoidance using an RGB-D camera. The proposed system first utilizes a depth image with an occupancy voxel map to generate potential dynamic obstacle regions as proposals. With the obstacle region proposals, the Kalman filter and our continuity filter are applied to track each dynamic obstacle. Finally, the environment-aware trajectory prediction method is proposed based on the Markov chain using the states of tracked dynamic obstacles. We implemented the proposed system with our custom quadcopter and navigation planner. The simulation and physical experiments show that our methods can successfully track and represent obstacles in dynamic environments in real-time and safely avoid obstacles.