Search

Agile quadrotor flight in challenging environments has the potential to revolutionize shipping, transportation, and search and rescue applications. Nonlinear model predictive control (NMPC) has recently shown promising results for agile quadrotor control, but relies on highly accurate models for maximum performance. Hence, model uncertainties in the form of unmodeled complex aerodynamic effects, varying payloads and parameter mismatch will degrade overall system performance. In this paper, we propose L1-NMPC, a novel hybrid adaptive NMPC to learn model uncertainties online and immediately compensate for them, drastically improving performance over the non-adaptive baseline with minimal computational overhead. Our proposed architecture generalizes to many different environments from which we evaluate wind, unknown payloads, and highly agile flight conditions. The proposed method demonstrates immense flexibility and robustness, with more than 90% tracking error reduction over non-adaptive NMPC under large unknown disturbances and without any gain tuning. In addition, the same controller with identical gains can accurately fly highly agile racing trajectories exhibiting top speeds of 70 km/h, offering tracking performance improvements of around 50% relative to the non-adaptive NMPC baseline., 8 Pages, 6 figures, Accepted RAL 2021

In this paper, we use a geometrical approach to sharpen a lower bound given in [5] for the Lipschitz modulus of the optimal value of (finite) linear programs under tilt perturbations of the objective function. The key geometrical idea comes from orthogonally projecting general balls on linear subspaces. Our new lower bound provides a computable expression for the exact modulus (as far as it only depends on the nominal data) in two important cases: when the feasible set has extreme points and when we deal with the Euclidean norm. In these two cases, we are able to compute or estimate the global Lipschitz modulus of the optimal value function in different perturbations frameworks.

Fault-tolerant control is crucial for safety-critical systems, such as quadrotors. State-of-art flight controllers can stabilize and control a quadrotor even when subjected to the complete loss of a rotor. However, these methods rely on external sensors, such as GPS or motion capture systems, for state estimation. To the best of our knowledge, this has not yet been achieved with only onboard sensors. In this paper, we propose the first algorithm that combines fault-tolerant control and onboard vision-based state estimation to achieve position control of a quadrotor subjected to complete failure of one rotor. Experimental validations show that our approach is able to accurately control the position of a quadrotor during a motor failure scenario, without the aid of any external sensors. The primary challenge to vision-based state estimation stems from the inevitable high-speed yaw rotation (over 20 rad/s) of the damaged quadrotor, causing motion blur to cameras, which is detrimental to visual inertial odometry (VIO). We compare two types of visual inputs to the vision-based state estimation algorithm: standard frames and events. Experimental results show the advantage of using an event camera especially in low light environments due to its inherent high dynamic range and high temporal resolution. We believe that our approach will render autonomous quadrotors safer in both GPS denied or degraded environments. We release both our controller and VIO algorithm open source., Comment: 8 pages, 10 figures

We present conditions for Holder calmness and upper Holder continuity of optimal solution sets to perturbed optimization problems in finite dimensions. Studies on Holder type stability were a popul...

Multirotor aerial robots are becoming widely used for the inspection of powerlines. To enable continuous, robust inspection without human intervention, the robots must be able to perch on the powerlines to recharge their batteries. Highly versatile perching capabilities are necessary to adapt to the variety of configurations and constraints that are present in real powerline systems. This paper presents a novel perching trajectory generation framework that computes perception-aware, collision-free, and dynamically-feasible maneuvers to guide the robot to the desired final state. Trajectory generation is achieved via solving a Nonlinear Programming problem using the Primal-Dual Interior Point method. The problem considers the full dynamic model of the robot down to its single rotor thrusts and minimizes the final pose and velocity errors while avoiding collisions and maximizing the visibility of the powerline during the maneuver. The generated maneuvers consider both the perching and the posterior recovery trajectories. The framework adopts costs and constraints defined by efficient mathematical representations of powerlines, enabling online onboard execution in resource-constrained hardware. The method is validated on-board an agile quadrotor conducting powerline inspection and various perching maneuvers with final pitch values of up to 180 degrees. The developed code is available online at: https://github.com/grvcPerception/pa_powerline_perching, IEEE Robotics and Automation Letters (2022)

Neural networks predictions are unreliable when the input sample is out of the training distribution or corrupted by noise. Being able to detect such failures automatically is fundamental to integrate deep learning algorithms into robotics. Current approaches for uncertainty estimation of neural networks require changes to the network and optimization process, typically ignore prior knowledge about the data, and tend to make over-simplifying assumptions which underestimate uncertainty. To address these limitations, we propose a novel framework for uncertainty estimation. Based on Bayesian belief networks and Monte-Carlo sampling, our framework not only fully models the different sources of prediction uncertainty, but also incorporates prior data information, e.g. sensor noise. We show theoretically that this gives us the ability to capture uncertainty better than existing methods. In addition, our framework has several desirable properties: (i) it is agnostic to the network architecture and task; (ii) it does not require changes in the optimization process; (iii) it can be applied to already trained architectures. We thoroughly validate the proposed framework through extensive experiments on both computer vision and control tasks, where we outperform previous methods by up to 23% in accuracy., Accepted for publication in the Robotics and Automation Letters 2020, and for presentation at the International Conference on Robotics and Automation (ICRA) 2020

Humans race drones faster than algorithms, despite being limited to a fixed camera angle, body rate control, and response latencies in the order of hundreds of milliseconds. A better understanding of the ability of human pilots of selecting appropriate motor commands from highly dynamic visual information may provide key insights for solving current challenges in vision-based autonomous navigation. This paper investigates the relationship between human eye movements, control behavior, and flight performance in a drone racing task. We collected a multimodal dataset from 21 experienced drone pilots using a highly realistic drone racing simulator, also used to recruit professional pilots. Our results show task-specific improvements in drone racing performance over time. In particular, we found that eye gaze tracks future waypoints (i.e., gates), with first fixations occurring on average 1.5 seconds and 16 meters before reaching the gate. Moreover, human pilots consistently looked at the inside of the future flight path for lateral (i.e., left and right turns) and vertical maneuvers (i.e., ascending and descending). Finally, we found a strong correlation between pilots eye movements and the commanded direction of quadrotor flight, with an average visual-motor response latency of 220 ms. These results highlight the importance of coordinated eye movements in human-piloted drone racing. We make our dataset publicly available., Comment: 8 pages, 6 figures

Autonomous car racing is a major challenge in robotics. It raises fundamental problems for classical approaches such as planning minimum-time trajectories under uncertain dynamics and controlling the car at the limits of its handling. Besides, the requirement of minimizing the lap time, which is a sparse objective, and the difficulty of collecting training data from human experts have also hindered researchers from directly applying learning-based approaches to solve the problem. In the present work, we propose a learning-based system for autonomous car racing by leveraging a high-fidelity physical car simulation, a course-progress proxy reward, and deep reinforcement learning. We deploy our system in Gran Turismo Sport, a world-leading car simulator known for its realistic physics simulation of different race cars and tracks, which is even used to recruit human race car drivers. Our trained policy achieves autonomous racing performance that goes beyond what had been achieved so far by the built-in AI, and, at the same time, outperforms the fastest driver in a dataset of over 50,000 human players., Comment: Accepted for Publication at the IEEE Robotics and Automation Letters (RA-L) 2021, and International Conference on Robots and Automation (ICRA) 2021

How do banks' interconnections in the euro area contribute to the vulnerability of the banking system? We study both the direct interconnections (banks lend to each other) and the indirect interconnections (banks are exposed to similar sectors of the economy). These complex linkages make the banking system more vulnerable to contagion risks., Comment les liens d’interdépendance entre les banques de la zone euro contribuent-ils à la vulnérabilité du système bancaire? Nous étudions les interconnexions directes (prêts entre les banques) et indirectes (exposition des banques à des secteurs semblables de l’économie). Ces liens complexes rendent le système bancaire plus vulnérable aux risques de contagion.

Recently, Canovas et al. presented an interesting result: the argmin mapping of a linear semi-infinite program under canonical perturbations is calm if and only if some associated linear semi-infinite inequality system is calm. Using classical tools from parametric optimization, we show that the if-direction of this condition holds in a much more general framework of optimization models, while the opposite direction may fail in the general case. In applications to special classes of problems, we apply a more recent result on the intersection of calm multifunctions.

During gait neurorehabilitation, many factors influence the quality of gait patterns, particularly the chosen body-weight support (BWS) device. Consequently, robotic BWS devices play a key role in gait rehabilitation of people with neurological disorders. The device transparency, support force vector direction, and attachment to the harness vary widely across existing robotic BWS devices, but the influence of these factors on the production of gait remains unknown. Because this information is key to designing an optimal BWS, we systematically studied these determinants in this work. We report that with a highly transparent device and a conventional harness, healthy participants select a small backward force when asked for optimal BWS conditions. This unexpected finding challenges the view that during human-robot interactions, humans predominantly optimize energy efficiency. Instead, they might seek to increase their feeling of stability and safety. We also demonstrate that the location of the attachment points on the harness strongly affects gait patterns, yet harness attachment is hardly reported in literature. Our results establish principles for the design of BWS devices and personalization of BWS settings for gait neurorehabilitation.

It is well known that there is a correspondence between convolutional codes and discrete-time linear systems over finite fields. In this paper, we employ the linear systems representation of a convolutional code to develop a decoding algorithm for convolutional codes over the erasure channel. In this kind of channel, which is important due to its use for data transmission over the Internet, the receiver knows if a received symbol is correct. We study the decoding problem using the state space description of a convolutional code, and this provides in a natural way additional information. With respect to previously known decoding algorithms, our new algorithm has the advantage that it is able to reduce the decoding delay as well as the computational effort in the erasure recovery process. We describe which properties a convolutional code should have in order to obtain a good decoding performance and illustrate it with an example.

In this letter, we study the effects that perception latency has on the maximum speed a robot can reach to safely navigate through an unknown cluttered environment. We provide a general analysis that can serve as a baseline for future quantitative reasoning for design tradeoffs in autonomous robot navigation. We consider the case where the robot is modeled as a linear second-order system with bounded input and navigates through static obstacles. Also, we focus on a scenario where the robot wants to reach a target destination in as little time as possible, and therefore cannot change its longitudinal velocity to avoid obstacles. We show how the maximum latency that the robot can tolerate to guarantee safety is related to the desired speed, the range of its sensing pipeline, and the actuation limitations of the platform (i.e., the maximum acceleration it can produce). As a particular case study, we compare monocular and stereo frame-based cameras against novel, low-latency sensors, such as event cameras, in the case of quadrotor flight. To validate our analysis, we conduct experiments on a quadrotor platform equipped with an event camera to detect and avoid obstacles thrown towards the robot. To the best of our knowledge, this is the first theoretical work in which perception and actuation limitations are jointly considered to study the performance of a robotic platform in high-speed navigation.

We study the determinants of convergence to efficient conventions in coordination games played on networks, when agents focus on past performance (imitative play). Previous theoretical results provide an incomplete picture and suggest potentially-complex interactions between the features of dynamics and behavior. We conducted an extensive simulation study (with approximately 1.12 million simulations) varying network size, interaction and information radius, the probability of actual interaction, the probability of mistakes, tie-breaking rules, and the process governing revision opportunities. Our main result is that “more interactions,” be it in the form of larger interaction neighborhoods or of a higher interaction probability, lead to less coordination on efficient conventions. A second observation, confirming previous but partial theoretical results, is that a large network size relative to the size of neighborhoods (a “large world”) facilitates convergence to efficient conventions. Third, a larger information neighborhood helps efficiency because it increases visibility of efficient payoffs across the network. Last, technical details of the dynamic specification as tie-breaking or inertia, while often relevant for specific theoretical results, appear to be of little empirical relevance in the larger space of dynamics.

Once an academic venture, autonomous driving has received unparalleled corporate funding in the last decade. Still, the operating conditions of current autonomous cars are mostly restricted to ideal scenarios. This means that driving in challenging illumination conditions such as night, sunrise, and sunset remains an open problem. In these cases, standard cameras are being pushed to their limits in terms of low light and high dynamic range performance. To address these challenges, we propose, DSEC, a new dataset that contains such demanding illumination conditions and provides a rich set of sensory data. DSEC offers data from a wide-baseline stereo setup of two color frame cameras and two high-resolution monochrome event cameras. In addition, we collect lidar data and RTK GPS measurements, both hardware synchronized with all camera data. One of the distinctive features of this dataset is the inclusion of high-resolution event cameras. Event cameras have received increasing attention for their high temporal resolution and high dynamic range performance. However, due to their novelty, event camera datasets in driving scenarios are rare. This work presents the first high-resolution, large-scale stereo dataset with event cameras. The dataset contains 53 sequences collected by driving in a variety of illumination conditions and provides ground truth disparity for the development and evaluation of event-based stereo algorithms., Comment: IEEE Robotics and Automation Letters

Event cameras are novel vision sensors that report per-pixel brightness changes as a stream of asynchronous "events". They offer significant advantages compared to standard cameras due to their high temporal resolution, high dynamic range and lack of motion blur. However, events only measure the varying component of the visual signal, which limits their ability to encode scene context. By contrast, standard cameras measure absolute intensity frames, which capture a much richer representation of the scene. Both sensors are thus complementary. However, due to the asynchronous nature of events, combining them with synchronous images remains challenging, especially for learning-based methods. This is because traditional recurrent neural networks (RNNs) are not designed for asynchronous and irregular data from additional sensors. To address this challenge, we introduce Recurrent Asynchronous Multimodal (RAM) networks, which generalize traditional RNNs to handle asynchronous and irregular data from multiple sensors. Inspired by traditional RNNs, RAM networks maintain a hidden state that is updated asynchronously and can be queried at any time to generate a prediction. We apply this novel architecture to monocular depth estimation with events and frames where we show an improvement over state-of-the-art methods by up to 30% in terms of mean absolute depth error. To enable further research on multimodal learning with events, we release EventScape, a new dataset with events, intensity frames, semantic labels, and depth maps recorded in the CARLA simulator.

Quadrotors are agile. Unlike most other machines, they can traverse extremely complex environments at high speeds. To date, only expert human pilots have been able to fully exploit their capabilities. Autonomous operation with on-board sensing and computation has been limited to low speeds. State-of-the-art methods generally separate the navigation problem into subtasks: sensing, mapping, and planning. While this approach has proven successful at low speeds, the separation it builds upon can be problematic for high-speed navigation in cluttered environments. Indeed, the subtasks are executed sequentially, leading to increased processing latency and a compounding of errors through the pipeline. Here we propose an end-to-end approach that can autonomously fly quadrotors through complex natural and man-made environments at high speeds, with purely onboard sensing and computation. The key principle is to directly map noisy sensory observations to collision-free trajectories in a receding-horizon fashion. This direct mapping drastically reduces processing latency and increases robustness to noisy and incomplete perception. The sensorimotor mapping is performed by a convolutional network that is trained exclusively in simulation via privileged learning: imitating an expert with access to privileged information. By simulating realistic sensor noise, our approach achieves zero-shot transfer from simulation to challenging real-world environments that were never experienced during training: dense forests, snow-covered terrain, derailed trains, and collapsed buildings. Our work demonstrates that end-to-end policies trained in simulation enable high-speed autonomous flight through challenging environments, outperforming traditional obstacle avoidance pipelines., Comment: 16 pages (+7 supplementary)

Reliable perception during fast motion maneuvers or in high dynamic range environments is crucial for robotic systems. Since event cameras are robust to these challenging conditions, they have great potential to increase the reliability of robot vision. However, event-based vision has been held back by the shortage of labeled datasets due to the novelty of event cameras. To overcome this drawback, we propose a task transfer method to train models directly with labeled images and unlabeled event data. Compared to previous approaches, (i) our method transfers from single images to events instead of high frame rate videos, and (ii) does not rely on paired sensor data. To achieve this, we leverage the generative event model to split event features into content and motion features. This split enables efficient matching between latent spaces for events and images, which is crucial for successful task transfer. Thus, our approach unlocks the vast amount of existing image datasets for the training of event-based neural networks. Our task transfer method consistently outperforms methods targeting Unsupervised Domain Adaptation for object detection by 0.26 mAP (increase by 93%) and classification by 2.7% accuracy.

Identifying risk spillovers in financial markets is of great importance for assessing systemic risk and portfolio management. Granger causality in tail (or in risk) tests whether past extreme events of a time series help predicting future extreme events of another time series. The topology and connectedness of networks built with Granger causality in tail can be used to measure systemic risk and to identify risk transmitters. Here we introduce a novel test of Granger causality in tail which adopts the likelihood ratio statistic and is based on the multivariate generalization of a discrete autoregressive process for binary time series describing the sequence of extreme events of the underlying price dynamics. The proposed test has very good size and power in finite samples, especially for large sample size, allows inferring the correct time scale at which the causal interaction takes place, and it is flexible enough for multivariate extension when more than two time series are considered in order to decrease false detections as spurious effect of neglected variables. An extensive simulation study shows the performances of the proposed method with a large variety of data generating processes and it introduces also the comparison with the test of Granger causality in tail by [Hong et al., 2009]. We report both advantages and drawbacks of the different approaches, pointing out some crucial aspects related to the false detections of Granger causality for tail events. An empirical application to high frequency data of a portfolio of US stocks highlights the merits of our novel approach., 22 pages, 7 Figures, 4 Tables

The Bitcoin system keeps its ledger consistent in a blockchain by solving cryptographic problems, in a method called "Proof-of-Work". The conventional wisdom asserts that the mining protocol is incentive-compatible. However, Eyal and Sirer in 2014 have discovered a mining attack strategy called selfish mining (SM), in which a miner (or a mining pool) publishes the blocks it mines selectively instead of immediately. SM strategy would have the impact of wasting resources of honest miners. Scholars proposed various extensions of the SM strategy and approaches to defense the SM attack. Whether selfish mining occurs in practice or not, has been subject of extensive debate. For the first time, in this paper we propose a method to identify selfish miners by detecting anomalies in the properties of consecutive blocks’ statistics. Furthermore, we extend our method to detect the mining cartels, in which miners secretly get together and share timely information. Our results provide evidence that these strategic behaviors take place in practice.

Today's autonomous drones have reaction times of tens of milliseconds, which is not enough for navigating fast in complex dynamic environments. To safely avoid fast moving objects, drones need low-latency sensors and algorithms. We departed from state-of-the-art approaches by using event cameras, which are bioinspired sensors with reaction times of microseconds. Our approach exploits the temporal information contained in the event stream to distinguish between static and dynamic objects and leverages a fast strategy to generate the motor commands necessary to avoid the approaching obstacles. Standard vision algorithms cannot be applied to event cameras because the output of these sensors is not images but a stream of asynchronous events that encode per-pixel intensity changes. Our resulting algorithm has an overall latency of only 3.5 milliseconds, which is sufficient for reliable detection and avoidance of fast-moving obstacles. We demonstrate the effectiveness of our approach on an autonomous quadrotor using only onboard sensing and computation. Our drone was capable of avoiding multiple obstacles of different sizes and shapes, at relative speeds up to 10 meters/second, both indoors and outdoors.

In this paper, we propose to augment image-based place recognition with structural cues. Specifically, these structural cues are obtained using structure-from-motion, such that no additional sensors are needed for place recognition. This is achieved by augmenting the 2D convolutional neural network (CNN) typically used for image-based place recognition with a 3D CNN that takes as input a voxel grid derived from the structure-from-motion point cloud. We evaluate different methods for fusing the 2D and 3D features and obtain best performance with global average pooling and simple concatenation. On the Oxford RobotCar dataset, the resulting descriptor exhibits superior recognition performance compared to descriptors extracted from only one of the input modalities, including state-of-the-art image-based descriptors. Especially at low descriptor dimensionalities, we outperform state-of-the-art descriptors by up to 90%., Comment: 8 pages, published in RA-L & IROS 2020

Motivated by the astonishing capabilities of natural intelligent agents and inspired by theories from psychology, this paper explores the idea that perception gets coupled to 3D properties of the world via interaction with the environment. Existing works for depth estimation require either massive amounts of annotated training data or some form of hard-coded geometrical constraint. This paper explores a new approach to learning depth perception requiring neither of those. Specifically, we train a specialized global-local network architecture with what would be available to a robot interacting with the environment: from extremely sparse depth measurements down to even a single pixel per image. From a pair of consecutive images, our proposed network outputs a latent representation of the observer's motion between the images and a dense depth map. Experiments on several datasets show that, when ground truth is available even for just one of the image pixels, the proposed network can learn monocular dense depth estimation up to 22.5% more accurately than state-of-the-art approaches. We believe that this work, despite its scientific interest, lays the foundations to learn depth from extremely sparse supervision, which can be valuable to all robotic systems acting under severe bandwidth or sensing constraints., Comment: Accepted for Publication at the IEEE Robotics and Automation Letters (RA-L) 2020, and International Conference on Intelligent Robots and Systems (IROS) 2020

Recent years have seen an increasing interest in the development of artificial intelligence circuits and systems for cloud-less edge computing applications. In an effort to reduce power consumption even further, we propose beyond von-Neumann in-memory computing architectures that can process the signals at the sensor side using ultra-low power mixed-signal analog/digital circuits which have properly matched dynamics and time-constants. In this paper, we propose one of the main computing elements of such architectures, namely the silicon neuron, designed using analog circuits in an advanced FDSOI 22 nm node. Here we optimize the design of an Adaptive Exponential Integrate and Fire (AdExp IF) neuron model for producing neural dynamics with biologically plausible time constants. We explore the options of the 22 nm FDSOI technology to address the analog design issues that arise from advanced scaling (such as leakage) and minimize power consumption by using a novel current comparator circuit with current-driven positive feedback. We present circuit simulation results which reproduce biologically plausible responses and compare the circuit energy per spike with state-of-the-art architectures. The proposed neuron design consumes one order of magnitude less power compared to the state-of-the-art and two orders of magnitude less compared to a pure digital implementation.

Among diagnostic biomarkers, high frequency oscillations in human iEEG are used to identify epileptogenic brain tissue during epilepsy surgery. However, current methods typically analyse the raw data offline using complex time-consuming algorithms. We developed a compact neuromorphic sensory-processing system-on-chip that can monitor the iEEG signals and detect high frequency oscillations in real-time using spiking neural networks. To this end, we present an integrated device with an analog front-end that can extract predefined spectral features and encode them as address-events, and a neuromorphic processor core that implements a network of integrate and fire neurons with dynamic synapses. The device was fabricated using a standard $0.18\mu\mathrm{m}$ CMOS technology node. The estimated power consumption of the front-end is $6.2\mu\mathrm{W}$ /channel and the area-on-chip for a single channel is 0.15 square millimetres. The SNN classifier provides 90.5% sensitivity and 67.7% specificity for detecting high frequency oscillations. This is the first feasibility study towards identifying relevant features in intracranial human data in real-time on-chip using event-base processors.

Continuous Integration (CI) is a widely-adopted software engineering practice. Despite its undisputed benefits, like higher software quality and improved developer productivity, mastering CI is not easy. Among the several barriers when transitioning to CI, developers need to face a new type of software failures (i.e., build failures) that requires them to understand complex build logs. Even when a team has successfully introduced a CI culture, living up to its principles and improving the CI practice are also challenging. In my research, I want to provide developers with the right support for establishing CI and the proper recommendations for continuously improving their CI process.

Civilian drones are soon expected to be used in a wide variety of tasks, such as aerial surveillance, delivery, or monitoring of existing architectures. Nevertheless, their deployment in urban environments has so far been limited. Indeed, in unstructured and highly dynamic scenarios, drones face numerous challenges to navigate autonomously in a feasible and safe way. In contrast to traditional “map-localize-plan” methods, this letter explores a data-driven approach to cope with the above challenges. To accomplish this, we propose DroNet: a convolutional neural network that can safely drive a drone through the streets of a city. Designed as a fast eight-layers residual network, DroNet produces two outputs for each single input image: A steering angle to keep the drone navigating while avoiding obstacles, and a collision probability to let the UAV recognize dangerous situations and promptly react to them. The challenge is however to collect enough data in an unstructured outdoor environment such as a city. Clearly, having an expert pilot providing training trajectories is not an option given the large amount of data required and, above all, the risk that it involves for other vehicles or pedestrians moving in the streets. Therefore, we propose to train a UAV from data collected by cars and bicycles, which, already integrated into the urban environment, would not endanger other vehicles and pedestrians. Although trained on city streets from the viewpoint of urban vehicles, the navigation policy learned by DroNet is highly generalizable. Indeed, it allows a UAV to successfully fly at relative high altitudes and even in indoor environments, such as parking lots and corridors. To share our findings with the robotics community, we publicly release all our datasets, code, and trained networks.

Best-reply behavior in Cournot oligopolies generally leads to Cournot-Nash equilibrium, but imitative behavior selects the Walrasian equilibrium as the unique stochastically stable state. Previous work (Alos-Ferrer, 2004) showed that in the presence of memory, imitative behavior leads to a non-trivial dynamics selecting all quantities between the Cournot and Walrasian outcomes. However, the scope of previous results was limited to specific assumptions on demand and cost functions, and did not provide information on the shape of the distribution of outcomes. We use computational simulations to address these limitations. We show that the selection result for non-trivial memory holds beyond the set of well-behaved Cournot games previously analyzed. Further, we find that, in Cournot games, the limit distribution of long-run outcomes is highly skewed towards the Walrasian quantity. Although longer memory increases the importance of the Cournot equilibrium, the competitive outcome remains the dominant prediction.

In recent years, many approaches to visual-inertial odometry (VIO) have become available. However, they neither exploit the robot's dynamics and known actuation inputs, nor differentiate between the desired motion due to actuation and the unwanted perturbation due to external force. For many robotic applications, it is often essential to sense the external force acting on the system due to, for example, interactions, contacts, and disturbances. Adding a motion constraint to an estimator leads to discrepancy between the model-predicted motion and the actual motion. Our approach exploits this discrepancy and resolves it by simultaneously estimating the motion and external force. We propose a relative motion constraint combining the robot's dynamics and the external force in a preintegrated residual, resulting in a tightly coupled, sliding-window estimator exploiting all correlations among all variables. We implement our visual inertial model-based odometry system into a state-of-the-art VIO approach and evaluate it against the original pipeline without motion constraints both on simulated and real-world data. The results show that our approach increases the accuracy of the estimator up to 29% as compared to the original VIO, and it provides external force estimates at no extra computational cost. To the best of our knowledge, this is the first approach that exploits model dynamics by jointly estimating motion and external force. Our implementation will be made available open-source.

A classical result due to Frank and Wolfe [An algorithm for quadratic programming. Naval Res Log Quart. 1956;3:95–110] says that a quadratic function f attains its supremum on a nonempty polyhedron M if f is bounded from above on M. In this note, we present a stringent proof of the extension of this result to cubic optimization (known from Andronov, Belousov and Shironin [On solvability of the problem of polynomial programming (In Russian). Izvestija Akadem. Nauk SSSR, Tekhnicheskaja Kibernetika. 1982;4:194–197. Translation appeared in News of the Academy of Science of USSR, Dept. of Technical Sciences, Technical Cybernetics.]). Further, we discuss related results. In particular, we bring back to attention Kummer's [Globale Stabilitat quadratischer Optimierungsprobleme. Wissenschaftliche Zeitschrift der Humboldt- Universitat zu Berlin, Math-Nat R. 1977;XXVI(5):565–569] generalization of the Frank-Wolfe theorem to the case that f is quadratic, but M is the Minkowski sum of a compact set and a polyh...

IEEE Robotics and Automation Letters, 4 (2), ISSN:2377-3766

The recent advances in state estimation, perception, and navigation algorithms have significantly contributed to the ubiquitous use of quadrotors for inspection, mapping, and aerial imaging. To further increase the versatility of quadrotors, recent works investigated the use of an adaptive morphology, which consists of modifying the shape of the vehicle during flight to suit a specific task or environment. However, these works either increase the complexity of the platform or decrease its controllability. In this letter, we propose a novel, simpler, yet effective morphing design for quadrotors consisting of a frame with four independently rotating arms that fold around the main frame. To guarantee stable flight at all times, we exploit an optimal control strategy that adapts on the fly to the drone morphology. We demonstrate the versatility of the proposed adaptive morphology in different tasks, such as negotiation of narrow gaps, close inspection of vertical surfaces, and object grasping and transportation. The experiments are performed on an actual, fully autonomous quadrotor relying solely on onboard visual-inertial sensors and compute. No external motion tracking systems and computers are used. This is the first work showing stable flight without requiring any symmetry of the morphology.

Making use of a structural model that allows for optimal liquidity management, we study the role that repos play in a bank׳s financing structure. In our model the bank׳s assets consist of illiquid loans and liquid reserves and are financed by a combination of repos, long-term debt, deposits and equity. Repos are a cheap source of funding, but they are subject to an exogenous rollover risk. We show that the use of repos inflicts two types of indirect (“shadow”) costs on the bank׳s shareholders: first, it induces the bank to maintain higher liquid reserves in order to alleviate the additional default risk; second, it adds to the cost of long-term debt financing. These shadow costs limit the bank׳s appetite for cheap but unstable repo funding. This effect is, however, weakened under poor returns on risky assets, access to deposit funding and the depositor preference rule. We also analyze the impact of a liquidity coverage ratio, payout restrictions and a leverage ratio on the bank׳s financing choices and show that all these tools are able to curb the bank׳s reliance on repos.

We investigate how the financial fragility in the real economy is affected by the average level of interdependence among agents across different regions of the economy. To this end, we develop a parsimonious agent-based model of firms and banks organized in geographic regions. The model is built on the framework of an existing class of models for business fluctuations. The goal of our exercise is to clarify the effect on systemic failures of the interplay between network interconnectedness and financial acceleration. In particular, we investigate the probability of individual and systemic failures with varying levels of interconnectedness. We find that, in the absence of financial acceleration, connectivity makes the system more resilient. In contrast, in the presence of financial acceleration, the probability of both individual and systemic failures are minimized at intermediate level of diversification.

This work proposes an IoT- and Blockchain-based, distributed system, for automated measuring, storing, and monitoring of water and air quality in environments such as lakes, mountains, urban areas, or factories. Comparable state-of-the-art solutions, require human interaction to access the data or require high power consumption or space requirements, or they are based on centralized architectures. The proposed pollution monitoring system here, on one hand, employs LoRa to address the high power consumption and long-range transmission challenges of IoT protocols. On the other hand, it is designed to be fully decentralized by using the Ethereum Blockchain to store and retrieve the data recorded by IoT sensors. Thus, data integrity is provided without the need for a Trusted Third Party (TTP) and data is collected and captured automatically without any manual operations needed. Observations on the four different types of sensors for measuring Potential Hydrogen (PH), Turbidity, Carbon monoxide (CO), and Carbon dioxide (CO2), revealed a high accuracy with the expected time-lines of measurements, non-falsified experimental values collected and can be used as reliable evidence of presence of pollution.

We use a model with agency frictions to analyze the structure of a dealer market that faces competition from a crossing network. Traders are privately informed about their types (e.g. their portfolios), which is something the dealer must take into account when engaging his counterparties. Instead of participating in the dealer market, the traders may take their business to a crossing network. The dealer must take into consideration that traders have this alternative when choosing a pricing schedule. We show that the presence of a crossing network may benefit traders even if they do not trade in it. Furthermore, it results in more traders being serviced by the dealer and the book’s spread shrinking (under certain conditions). We allow for the pricing on the dealer market to determine the structure of the crossing network, which itself influences the structure of the dealer market. This results in a feedback look that, under the same conditions that lead to a reduction of the spread, yields an equilibrium book/crossing network pair.

We examine how agents allocate attention between private and public signals to reduce the uncertainty about observation noises when coordination is an important concern. In this setting, the attention allocation may not be monotone in endowed attention capacity. Agents may decrease their attention on or even ignore the more accurate signal when capacity increases. As a result, social welfare may decrease when they have more attention to process information. And it can be even higher when agents possess a finite amount of capacity than when they have an infinite amount of capacity. We derive sufficient and necessary conditions under which multiple equilibria emerge and study the implications of equilibrium multiplicity for macroeconomic policies.

To what extent is public debt private liquidity? Much policy advice given in the aftermath of the financial crisis rests on the assumption that increasing public debt relaxes borrowing constraints of private households. This is the case for ad-hoc debt limits, which are exogenous to public policy. Instead, if debt limits are fully endogenous, as e.g. in the case of the natural borrowing limit, public debt has no impact. We assume that borrowing limits arise because of limited contract enforceability and are therefore determined as equilibrium outcomes. Using an incomplete markets economy in which households are subject to uninsurable earnings shocks, we show that public debt provides some liquidity, but less so than it would if constraints were imposed ad-hoc. We show that generating borrowing constraints as an equilibrium outcome substantially alters the answers to other important questions, such as for the welfare effects of government debt or its impact on real economic activity.

We present an algorithm to estimate the rotational motion of an event camera. In contrast to traditional cameras, which produce images at a fixed rate, event cameras have independent pixels that respond asynchronously to brightness changes, with microsecond resolution. Our method leverages the type of information conveyed by these novel sensors (i.e., edges) to directly estimate the angular velocity of the camera, without requiring optical flow or image intensity estimation. The core of the method is a contrast maximization design. The method performs favorably against ground truth data and gyroscopic measurements from an Inertial Measurement Unit, even in the presence of very high-speed motions (close to 1000 deg/s).

Unlike standard cameras that send intensity images at a constant frame rate, event-driven cameras asynchronously report pixel-level brightness changes, offering low latency and high temporal resolution (both in the order of micro-seconds). As such, they have great potential for fast and low power vision algorithms for robots. Visual tracking, for example, is easily achieved even for very fast stimuli, as only moving objects cause brightness changes. However, cameras mounted on a moving robot are typically non-stationary and the same tracking problem becomes confounded by background clutter events due to the robot ego-motion. In this paper, we propose a method for segmenting the motion of an independently moving object for event-driven cameras. Our method detects and tracks corners in the event stream and learns the statistics of their motion as a function of the robot's joint velocities when no independently moving objects are present. During robot operation, independently moving objects are identified by discrepancies between the predicted corner velocities from ego-motion and the measured corner velocities. We validate the algorithm on data collected from the neuromorphic iCub robot. We achieve a precision of ~ 90 % and show that the method is robust to changes in speed of both the head and the target., Comment: 7 pages, 6 figures

This paper contributes to a growing literature on the ambiguous effects of risk diversification. In our model, banks hold claims on each other’s liabilities that are marked-to-market on the individual financial leverage of the obligor. The probability of systemic default is determined using a passage-problem approach in a network context and banks are able to internalize the network externalities of contagion through their holdings. Banks do not internalize the social costs to the real economy of a systemic default of the banking system. We investigate the optimal diversification strategy of banks in the face of opposite and persistent economic trends that are ex-ante unknown to banks. We find that the optimal level of risk diversification may be interior or extremal depending on banks exposure the external assets and that a tension arises whereby individual incentives favor a banking system that is over-diversified with respect to the level of diversification that is desirable in the social optimum.

We calculate the probability that random polynomial matrices over a finite field with certain structures are right prime or left prime, respectively. In particular, we give an asymptotic formula for the probability that finitely many non-singular polynomial matrices are mutually left coprime. These results are used to estimate the number of reachable and observable linear systems as well as the number of non-catastrophic convolutional codes. Moreover, we are able to achieve an asymptotic formula for the probability that a parallel connected linear system is reachable.

This paper describes the application of a Convolutional Neural Network (CNN) in the context of a predator/prey scenario. The CNN is trained and run on data from a Dynamic and Active Pixel Sensor (DAVIS) mounted on a Summit XL robot (the predator), which follows another one (the prey). The CNN is driven by both conventional image frames and dynamic vision sensor "frames" that consist of a constant number of DAVIS ON and OFF events. The network is thus "data driven" at a sample rate proportional to the scene activity, so the effective sample rate varies from 15 Hz to 240 Hz depending on the robot speeds. The network generates four outputs: steer right, left, center and non-visible. After off-line training on labeled data, the network is imported on the on-board Summit XL robot which runs jAER and receives steering directions in real time. Successful results on closed-loop trials, with accuracies up to 87% or 92% (depending on evaluation criteria) are reported. Although the proposed approach discards the precise DAVIS event timing, it offers the significant advantage of compatibility with conventional deep learning technology without giving up the advantage of data-driven computing., Paper presented at the conference: Second International Conference on Event-Based Control, Communication and Signal Processing (EBCCSP) 2016, At Krakow, Poland