Search

Flying insects employ elegant optical-flow-based strategies to solve complex tasks such as landing or obstacle avoidance. Roboticists have mimicked these strategies on flying robots with only limited success, because optical flow (1) cannot disentangle distance from velocity and (2) is less informative in the highly important flight direction. Here, we propose a solution to these fundamental shortcomings by having robots learn to estimate distances to objects by their visual appearance. The learning process obtains supervised targets from a stability-based distance estimation approach. We have successfully implemented the process on a small flying robot. For the task of landing, it results in faster, smooth landings. For the task of obstacle avoidance, it results in higher success rates at higher flight speeds. Our results yield improved robotic visual navigation capabilities and lead to a novel hypothesis on insect intelligence: behaviours that were described as optical-flow-based and hardwired actually benefit from learning processes. Autonomous flight is challenging for small flying robots, given the limited space for sensors and on-board processing capabilities, but a promising approach is to mimic optical-flow-based strategies of flying insects. A new development improves this technique, enabling smoother landings and better obstacle avoidance, by giving robots the ability to learn to estimate distances to objects by their visual appearance.

ROVIO is one of the state-of-the-art monocular visual inertial odometry algorithms. It uses an Iterative Extended Kalman Filter (IEKF) to align visual features and update the vehicle state simultaneously by including the feature locations in the state vector of the IEKF. This algorithm is single-core intensive, which allows the other cores to be used for other algorithms, such as object detection and path optimization. However, the computational cost of the algorithm grows rapidly with the maximum number of features to track. Each feature adds three new states (a 2D bearing vector and inverse depth), leading to bigger matrix multiplications that are computationally expensive. The main computational load of ROVIO is the iterative update step of the IEKF. In this work, we reduce the average computational cost of ROVIO by [Formula: see text] on an NVIDIA Jetson TX2, without affecting the accuracy of the algorithm. This computational gain is mainly achieved by exploiting the sparse matrices in ROVIO. Furthermore, we reduce the computational peaks by pre-selecting new features based on their already calculated FAST score. The combination of both modifications allows us to run ROVIO on the computationally restricted Raspberry Pi Zero 2[Formula: see text]W.

A lot of UAV applications require vertical take-off and landing (VTOL) combined with very long-range or endurance. Transitioning UAVs have been proposed to combine the VTOL capabilities of helicopters with the efficient long-range flight properties of fixed-wing aircraft. But energy is still a bottleneck for many electric long endurance applications. While solar power technology and battery technology have improved a lot, in rougher conditions they still respectively lack the power or total amount of energy required for many real-world situations. In this paper, we introduce the NederDrone, a hybrid lift, hybrid energy hydrogen-powered UAV which can perform vertical take-off and landings using 12 propellers while flying efficiently in forward flight thanks to its fixed wings. The energy is supplied from a mix of hydrogen-driven fuel-cells to store large amounts of energy and battery power for high power situations. The hydrogen is stored in a pressurized cylinder around which the UAV is optimized. This paper analyses the selection of the concept, the implemented safety elements, the electronics and flight control and shows flight data including a 3h38 flight at sea, starting and landing on a small moving ship.

Purpose: The study was performed comparing dosimetric characteristics of volumetric modulated arc therapy (VMAT) and field-in-field (FiF) techniques on a patient with synchronous bilateral breast carcinoma. Methods: The patients with bilateral breast cancer treatment were included in this study. A total dose of 40.05 Gy in 15 fractions was prescribed to the Planning Target Volume (PTV) of the whole bilateral breast cancer with the supraclavicular and infraclavicular nodes, with a complementary boost of 10 Gy in 4 fractions to the surgical bed (PTVboost). For both radiotherapy techniques, several VxGy parameters were analyzed for the PTVs, together with the Conformity index (CI), the Homogeneity index (HI) and the critical organs at risk (OARs), lungs and heart. Results: The patient was treated by the VMAT technique and the daily treatment time was less than 20 minutes with daily CBCT imaging. In the VMAT plan, the PTV 95% dose covered 38.89 ± 0.81 Gy, compared to 37.26 ± 1.02 Gy in the FiF technique. The VMAT plan improved the dose homogeneity index and lower dose in lung towards high dose region. Conclusion: The study demonstrates the viability of the VMAT technique in the treatment of bilateral breast cancer. The introduced single isocentric VMAT technique is fast to deliver and it increases the dose homogeneity of the target volume with some limitations. The treatment was well tolerated, without interruption of the treatment courses caused by treatment-related toxicities.

Hybrid UAVs have gained a lot of interest for their combined vertical take-off & landing (VTOL) and efficient forward flight capabilities. But their control is facing challenges in over-actuation and conflicting requirements depending on the flight phase which can easily lead to actuator saturation. Incremental Non-linear Control Allocation (INCA) has been proposed to solve the platform’s control allocation problem in the case of saturation or over-actuation by minimizing a set of objective functions. This work demonstrates INCA on quadplanes, an in-plane combination between a quadrotor and a conventional fixed-wing, and proposes an extension to control the outer loop. The novel controller is called Extended INCA (XINCA) and adds the wing orientation as a force-generating actuator in the outerloop control optimization. This leads to a single controller for all flight phases that avoids placing the wing at negative angles of attack and minimizes the load on hover motors. XINCA has low dependence on accurate vehicle models and requires only several optimization parameters. Flight simulations and experimental flights are performed to demonstrate the performance.

PurposeIn advanced radiotherapy techniques such as intensity-modulated radiation therapy (IMRT), the quality assurance (QA) process is essential. The aim of the study was to assure the treatment planning dose delivered during delivery of complex treatment plans. The QA standard is to perform patient-specific comparisons between planned doses and doses measured in a phantom.Materials and methodThe Delta 4 phantom (Scandidos, Uppsala, Sweden) has been used in this study. This device consists of diode matrices in two orthogonal planes inserted in a cylindrical acrylic phantom. Each diode is sampled per beam pulse so that the dose distribution can be evaluated on segment-by-segment, beam-by-beam, or as a composite plan from a single set of measurements. Ninety-five simple and complex radiotherapy treatment plans for different pathologies, planned using a treatment planning system (TPS) were delivered to the QA device. The planned and measured dose distributions were then compared and analysed. The gamma index was determined for different pathologies.ResultsThe evaluation was performed in terms of dose deviation, distance to agreement and gamma index passing rate. The measurements were in excellent agreement between with the calculated dose of the TPS and the QA device. Overall, good agreement was observed between measured and calculated doses in most cases with gamma values above 1 in >95% of measured points. Plan results for each test met the recommended dose goals.ConclusionThe delivery of IMRT and volumetric-modulated arc therapy (VMAT) plans was verified to correspond well with calculated dose distributions for different pathologies. We found the Delta 4 device is accurate and reproducible. Although Delta4 appears to be a straightforward device for measuring dose and allows measure in real-time dosimetry QA, it is a complex device and careful quality control is required before its use.

Monocular optical flow has been widely used to detect obstacles in Micro Air Vehicles (MAVs) during visual navigation. However, this approach requires significant movement, which reduces the efficiency of navigation and may even introduce risks in narrow spaces. In this paper, we introduce a novel setup of self-supervised learning (SSL), in which optical flow cues serve as a scaffold to learn the visual appearance of obstacles in the environment. We apply it to a landing task, in which initially 'surface roughness' is estimated from the optical flow field in order to detect obstacles. Subsequently, a linear regression function is learned that maps appearance features represented by texton distributions to the roughness estimate. After learning, the MAV can detect obstacles by just analyzing a still image. This allows the MAV to search for a landing spot without moving. We first demonstrate this principle to work with offline tests involving images captured from an on-board camera, and then demonstrate the principle in flight. Although surface roughness is a property of the entire flow field in the global image, the appearance learning even allows for the pixel-wise segmentation of obstacles., This manuscript is submitted to Robotics and Autonomous Systems

In the AlphaPilot Challenge, teams compete to fly autonomous drones through an obstacle course as fast as possible. The 2019 winning team MAVLab reflects on the challenge of beating human pilots.

Swarms of tiny flying robots hold great potential for exploring unknown, indoor environments. Their small size allows them to move in narrow spaces, and their light weight makes them safe for operating around humans. Until now, this task has been out of reach due to the lack of adequate navigation strategies. The absence of external infrastructure implies that any positioning attempts must be performed by the robots themselves. State-of-the-art solutions, such as simultaneous localization and mapping, are still too resource demanding. This article presents the swarm gradient bug algorithm (SGBA), a minimal navigation solution that allows a swarm of tiny flying robots to autonomously explore an unknown environment and subsequently come back to the departure point. SGBA maximizes coverage by having robots travel in different directions away from the departure point. The robots navigate the environment and deal with static obstacles on the fly by means of visual odometry and wall-following behaviors. Moreover, they communicate with each other to avoid collisions and maximize search efficiency. To come back to the departure point, the robots perform a gradient search toward a home beacon. We studied the collective aspects of SGBA, demonstrating that it allows a group of 33-g commercial off-the-shelf quadrotors to successfully explore a real-world environment. The application potential is illustrated by a proof-of-concept search-and-rescue mission in which the robots captured images to find "victims" in an office environment. The developed algorithms generalize to other robot types and lay the basis for tackling other similarly complex missions with robot swarms in the future.

To investigate how an unmanned air vehicle can detect manned aircraft with a single microphone, an audio data set is created in which unmanned air vehicle ego-sound and recorded aircraft sound are mixed together. A convolutional neural network is used to perform air traffic detection. Due to restrictions on flying unmanned air vehicles close to aircraft, the data set has to be artificially produced, so the unmanned air vehicle sound is captured separately from the aircraft sound. They are then mixed with unmanned air vehicle recordings, during which labels are given indicating whether the mixed recording contains aircraft audio or not. The model is a convolutional neural network that uses the features Mel frequency cepstral coefficient, spectrogram or Mel spectrogram as input. For each feature, the effect of unmanned air vehicle/aircraft amplitude ratio, the type of labeling, the window length and the addition of third party aircraft sound database recordings are explored. The results show that the best performance is achieved using the Mel spectrogram feature. The performance increases when the unmanned air vehicle/aircraft amplitude ratio is decreased, when the time window is increased or when the data set is extended with aircraft audio recordings from a third party sound database. Although the currently presented approach has a number of false positives and false negatives that is still too high for real-world application, this study indicates multiple paths forward that can lead to an interesting performance. Finally, the data set is provided as open access.

Autonomous flight of Flapping Wing Micro Air Vehicles (FWMAVs) is a major challenge in the field of robotics, due to their light weight and their flapping-induced body motions. An FWMAV is presented weighing a mere 20 g while all its sensors and processing for autonomous flight are onboard. The navigation is based on a 4-g stereo vision camera with onboard processing. Three basic navigational tasks are demonstrated, namely obstacle avoidance, door traversing and corridor following. The presented combination of sensors and control routines is shown to allow flight in common unprepared environments like corridors and offices. The algorithms do not depend on prior classification or learning of the environment or control logic and work in any unprepared environment with vertical texture. While some failure cases remain, this work forms an important step towards very small autonomous indoor MAV.

This paper presents an approach to the system identification of the Delfly II Flapping Wing Micro Air Vehicle (FWMAV) using flight test data. It aims at providing simple FWMAV aerodynamic models that can be used in simulations as well as in nonlinear flight control systems. The undertaken methodology builds on normal aircraft system identification methods and extends these with techniques that are specific to FWMAV model identification. The entire aircraft model identification cycle is discussed covering the set-up and automatic execution of the flight test experiments, the aircraft states, the aerodynamic forces and moments' reconstruction, the aerodynamic model structure selection, the parameter estimation and finally, the model validation. In particular, a motion capturing facility was used to record the flapper's position in time and from there compute the states and aerodynamic forces and moments that acted on it, assuming flap-averaged dynamics and linear aerodynamic model structures. It is shown that the approach leads to aerodynamic models that can predict the aerodynamic forces with high accuracy. Despite less accurate, the predictions of the aerodynamic moments still follow the general trend of the measured moments. Dynamic simulations based on the identified aerodynamic models show flight trajectories that closely match the ground truth spanning a number of flapping cycles. Finally, the dimensional aerodynamic forces and moments' coefficients of two of the identified aerodynamic models are presented.

This work extends the drag-force enhanced quadrotor model by denoting the free stream air velocity as the difference between the ground speed and the wind speed. It is demonstrated that a relatively simple nonlinear observer is capable of estimating the local wind components, provided accelerometer and GPS-velocity measurements are available. We perform a wind tunnel experiment at various wind speeds using a quadrotor vehicle with a low-cost Inertial Measurement Unit (IMU) and a motion tracking system to provide accurate ground speed measurements. It is shown that the onboard Extended Kalman Filter (EKF) accurately estimates the wind components.

Purpose To assess the dosimetric impact of a patient positioning device for prone breast radiotherapy and assess the accuracy of a treatment planning system (TPS) in predicting this impact. Methods Beam attenuation and build-up dose perturbations, quantified by ionization chamber and radiochromic film dosimetry, were evaluated for 3 components of the patient positioning device: the carbon fiber baseplate, the support cushions and the support wedge for the contralateral breast. Dose calculations were performed using the XVMC dose engine implemented in the Monaco TPS. All components were included during planning CT acquisition. Results Beam attenuation amounted to 7.57% (6 MV) and 5.33% (15 MV) for beams obliquely intersecting the couchtop–baseplate combination. Beams traversing large sections of the support wedge were attenuated by 12.28% (6 MV) and 9.37% (15 MV). For the support cushion foam, beam attenuation remained limited to 0.11% (6 MV) and 0.08% (15 MV) per centimeter thickness. A substantial loss of dose build-up was detected when irradiating through any of the investigated components. TPS dose calculations accurately predicted beam attenuation by the baseplate and support wedge. A manual density overwrite was needed to model attenuation by the support cushion foam. TPS dose calculations in build-up regions differed considerably from measurements for both open beams and beams traversing the device components. Conclusions Irradiating through the components of the positioning device resulted in a considerable degradation of skin sparing. Inclusion of the device components in the treatment planning CT allowed to accurately model the most important attenuation effect, but failed to accurately predict build-up doses.

The radiochromic film, which is used, in combination with a flatbed scanner has become a widely used tool for a quantitative evaluation of radiation dose in radiation therapy. One aspect of uncertainty using the radiochromic film is the magnitude of orientation effects when the orientation of the film is not kept constant during the digitization process. The aim of this note was to investigate the impact of using a combination of two crossed sheets of EBT2 film on various aspects of radiochromic film dosimetry. First the impact on the film sensitivity was studied. We also investigated the influence on orientation effects during scanning. The results show that the double crossed film combination increases the sensitivity with a factor 1.7-2.1 and practically eliminates the effects of film orientation on the optical density read-out and the lateral correction profiles.

Purpose: The dosimetric accuracy of EDR2 radiographic film has been rigorously assessed in regular and intensity modulated beams for various incidence angles, including the parallel and perpendicular orientation. There clearly exists confusion in literature regarding the effect of film orientation. The primary aim is to clarify potential sources of the confusion and to gain physical insight into the film orientation effect with a link to radiochromic film as well. Methods: An inverse pyramid IMRT field, consisting of six regular and elongated 3 Multiplication-Sign 20 cm{sup 2} field segments, was studied in perpendicular and parallel orientation. Assessment of film self-perturbation and intrinsic directional sensitivity were also included in the experiments. Finally, the authors investigated the orientational effect in composite beams in the two extreme orientations, i.e., perpendicular and parallel. Results: The study of an inverse pyramid dose profile revealed good agreement between the perpendicular film and the diamond detector within 0.5% in the low-scatter regions for both 6 and 18 MV. The parallel oriented film demonstrated a 3% under-response at 5-cm (6 MV) depth against the perpendicular orientation, but both orientations over responded equally in the central region, which received only scattered dose, at both 5- and 20-cm depths. In amore » regular 6-MV 5 Multiplication-Sign 5 cm{sup 2} field, a 4.1% lower film response was observed in the parallel orientation compared to perpendicular orientation. The under response gradually increased to 6% when reducing the field size to 0.5 Multiplication-Sign 5 cm{sup 2}. On the other hand, the film showed a 1.7% lower response in parallel orientation for the large field size of 20 Multiplication-Sign 20 cm{sup 2} at 5-cm depth but the difference disappeared at 10 cm. At 18 MV, similar but somewhat lower differences were found between the two orientations. The directional sensitivity of the film diminishes with increasing field size and depth. Surprisingly a composite IMRT beam consisting of 20 adjacent strip segments also produced a significant orientational dependence of film response, notwithstanding the large total field size of 20 Multiplication-Sign 20 cm{sup 2}. Conclusions: This analysis allowed the development of a hypothesis about the physics behind the orientational dependence of film response in general and to formulate precautions when using film dosimetry in the dosimetric verification of multibeam treatments.« less

Superficial dose from 6- and 18-MV photon beams has been studied by measuring surface dose and shallow build-up dose using radiographic film EDR2, radiochromic film EBT2 and plane-parallel chamber. Measurements have been made for intensity- and non-intensity-modulated beams. The results show that the surface dose was found to be 19.8% and 10% of maximum dose in unmodulated fields for 6 and 18 MV photon beams, respectively. The study further showed that intensity modulation decreased surface dose by 1.1% and 0.7% for the same field size at 6 and 18 MV, respectively, and surface dose was dropped by magnetically sweeping contaminating electrons. EDR2 and EBT2 films show in good agreement in shallow build-up region. This study demonstrated the capability of EDR2 film, in addition to radiochromic film, to measure surface and build-up dose in case of treatment planning system uncertainties with regard to skin toxicity or shallow target coverage.

Purpose: To evaluate the dose distribution of a 120-MeV laser-plasma accelerated electron beam which may be of potential interest for high-energy electron radiation therapy. Methods: In the interaction between an intense laser pulse and a helium gas jet, a well collimated electron beam with very high energy is produced. A secondary laser beam is used to optically control and to tune the electron beam energy and charge. The potential use of this beam for radiation treatment is evaluated experimentally by measurements of dose deposition in a polystyrene phantom. The results are compared to Monte Carlo simulations using the GEANT4 code. Results: It has been shown that the laser-plasma accelerated electron beam can deliver a peak dose of more than 1 Gy at the entrance of the phantom in a single laser shot by direct irradiation, without the use of intermediate magnetic transport or focusing. The dose distribution is peaked on axis, with narrow lateral penumbra. Monte Carlo simulations of electron beam propagation and dose deposition indicate that the propagation of the intense electron beam (with large self-fields) can be described by standard models that exclude collective effects in the response of the material. Conclusions: The measurements show that the high-energy electron beams produced by an optically injected laser-plasma accelerator can deliver high enough dose at penetration depths of interest for electron beam radiotherapy of deep-seated tumors. Many engineering issues must be resolved before laser-accelerated electrons can be used for cancer therapy, but they also represent exciting challenges for future research. (C) 2012 American Association of Physicists in Medicine. [http://dx.doi.org/10.1118/1.47199621 (Less)

The appearance variation cue captures the variation in texture in a single image. Its use for obstacle avoidance is based on the assumption that there is less such variation when the camera is close to an obstacle. For videos of approaching frontal obstacles, it is demonstrated that combining the cue with optic flow leads to better performance than using either cue alone. In addition, the cue is successfully used to control the 16-g flapping-wing micro air vehicle DelFly II.

The aim of this study is to determine the optimal backscatter thickness and lateral phantom dimension beyond the irradiated volume for the dosimetric verification with radiographic film when applying large field sizes. Polystyrene and Virtual Water™ phantoms were used to study the influence of the phantom backscatter thickness. EDR2 and XV films were used in 6 and 18 MV photon beams. The results show 11.4% and 6.4% over-response of the XV2 film when compared to the ion chamber for 6 MV 30×30 and 10×10 cm2 field sizes, respectively, when the phantom backscatter thickness is 5 cm. For the same setup, measurements with EDR2 films indicate 8.5% and 1.7% over-response. The XV2 film response in the polystyrene phantom is about 2.0% higher than in the Virtual Water™ phantom for the 6 MV beam and 20 cm backscatter thickness. Similar results were obtained for EDR2 film. In the lateral scatter study, film response was nearly constant within 5 cm of lateral thickness and it increases when lateral thickness increases due to more multiple scatter of low energy photons. The backscatter thickness of the phantom should be kept below 7 cm for the accuracy of the film dosimetry. The lateral extension of the phantom should not be more than 5 cm from the field boundary in case of large irradiated volumes.

Small robotic systems such as Micro Air Vehicles (MAVs) need to react quickly to their dynamic environments, while having only a limited amount of energy and processing onboard. In this article, sub-sampling of local image samples is investigated as a straightforward and broadly applicable approach to improve the computational efficiency of vision algorithms. In sub-sampling, only a small subset of the total number of samples is processed, leading to a significant reduction of the computational effort at the cost of a slightly lower accuracy. The possibility to change the number of extracted samples is of particular importance to autonomous robots, since it allows the designer to select not only the performance but also the execution frequency of the algorithm. The approach of sub-sampling is illustrated by introducing two novel, computationally efficient algorithms for two tasks relevant to MAVs: WiFi noise detection in camera images and onboard horizon detection for pitch and roll estimation. In the noise detection task, image lines and pixel pairs are sampled, while in the horizon detection task features from local image patches are sampled. For both tasks experiments are performed and the effects of sub-sampling are analyzed. It is demonstrated that even for small images of size 160x120 speed-ups of a factor 14 to 21 are reached, while retaining a sufficient performance for the tasks at hand.

Light-weight, autonomous ornithopters form a promise to observe places that are too small or too dangerous for humans to enter. In this article, we discuss the DelFly project, in which we follow a top-down approach to ever smaller and more autonomous ornithopters. Top-down signifies that the project always focuses on complete flying systems equipped with camera. We give arguments for the approach by explaining which findings on the DelFly I and DelFly II recently led to the development of the DelFly Micro: a 3.07-gram ornithopter carrying a camera and transmitter onboard. These findings concern the design, aerodynamics, and vision-based control of the DelFly. In addition, we identify main obstacles on the road to fly-sized ornithopters.

The purpose of this investigation was the verification of both the measured data and quality of the implementation of the add-on ModuLeaf miniature multileaf collimator (ML mMLC) into the clinical treatment planning system for conformal stereotactic radiosurgery treatment. To this end the treatment head with ML mMLC was modeled in the BEAMnrc Monte Carlo (MC) code. The 6 MV photon beams used in the setup were first benchmarked with a set of measurements. A total ML mMLC transmission of 1.13% of the 10 x 10 cm2 open field dose was measured and reproduced with the BEAMnrc/DOSXYZnrc code. Correspondence between calculated and measured output factors (OFs) was within 2%. Correspondence between MC and measured profiles was within 2% dose and 2 mm distance, only for the smallest 0.5 x 0.5 cm2 field the results were within 3% dose. In the next step, the MC model was compared with Gafchromic film measurements and Pinnacle(3) 7.4 f (convolution superposition algorithm) calculated dose distributions, using a gamma evaluation comparison, for a multi-beam patient setup delivered to a Lucytrade mark phantom. The gamma evaluation of the MC versus Gafchromic film resulted in 3.4% of points not fulfilling gamma

During the last few decades, accuracy in photon and electron radiotherapy has increased substantially. This is partly due to enhanced linear accelerator technology, providing more flexibility in field definition (e.g. the usage of computer-controlled dynamic multileaf collimators), which led to intensity modulated radiotherapy (IMRT). Important improvements have also been made in the treatment planning process, more specifically in the dose calculations. Originally, dose calculations relied heavily on analytic, semi-analytic and empirical algorithms. The more accurate convolution/superposition codes use pre-calculated Monte Carlo dose “kernels” partly accounting for tissue density heterogeneities. It is generally recognized that the Monte Carlo method is able to increase accuracy even further. Since the second half of the 1990s, several Monte Carlo dose engines for radiotherapy treatment planning have been introduced. To enable the use of a Monte Carlo treatment planning (MCTP) dose engine in clinical circumstances, approximations have been introduced to limit the calculation time. In this paper, the literature on MCTP is reviewed, focussing on patient modeling, approximations in linear accelerator modeling and variance reduction techniques. An overview of published comparisons between MC dose engines and conventional dose calculations is provided for phantom studies and clinical examples, evaluating the added value of MCTP in the clinic. An overview of existing Monte Carlo dose engines and commercial MCTP systems is presented and some specific issues concerning the commissioning of a MCTP system are discussed.

In this chapter, we first explain the choice for the X-wing configuration of the DelFly. Subsequently, we discuss the evolution of the different mechanical parts used in the DelFly I, II, Micro, and Explorer. Details are given on the used materials.

In this chapter, an overview of the experimental studies focusing on the aerodynamics of the DelFly is given. Force generation mechanisms of the DelFly flapping-wing flight are addressed with particular consideration of the effects of different parameters such as wing geometry and flexibility. Furthermore, state-of-art flow visualization results are presented to give insight into the behavior of flow structures around and in the wake of the flapping wings and to assess their influence on the unsteady forces.

This chapter treats the main choices, issues, and tradeoffs in the design of flapping wing MAVs. In particular, we discuss the implications of different tail and wing configurations, the energy source and various types of actuators. We also show how choices elementary to aircraft design, such as the trade-off between fuel/battery mass and payload mass can have rather large effects at the scale of light-weight flapping wing MAVs.

The DelFly does not use a map of its environment to achieve successful obstacle avoidance. Instead, it uses more direct visual cues to detect obstacles and to turn in the right direction. In this chapter, we investigate how successful such a direct avoidance strategy can be. We introduce a turning logic for the DelFly and test it both in simulation and on the real DelFly II platform, confirming the validity of the approach and identifying some of the challenges still ahead.

We shift our focus to the use of stereo vision for autonomous flight. Stereo vision implies carrying two cameras on board, which adds weight and increases the power consumption. Still, it also allows for instantaneous distance estimates, which is a considerable advantage on a moving (and oscillating) flapping wing MAV. In particular, we explain the onboard stereo vision and control algorithms that allow the 20-g DelFly Explorer to autonomously fly around in unknown environments for as long as its battery lasts.

In this chapter, physical principles of fixed-wing and flapping-wing aerodynamics are introduced briefly with the associated terminology. Furthermore, the main aerodynamic mechanisms which are responsible for the generation of the forces in the flapping-wing flight are described in order to provide the reader with fundamental knowledge before further comprehensive analysis of the DelFly.

This chapter sets the stage for the research on autonomous flight of the DelFly. First, a general introduction is given to artificial intelligence for robotics. This will permit the layman to understand some of the major challenges in creating autonomous robots, and the different solution approaches. Subsequently, the particular tasks and approaches to autonomous flight of Micro Air Vehicles are discussed. Whereas mainstream approaches are now being applied to \(\sim \)1 kg MAVs such as quadrotors, it will become clear that light-weight flapping wing MAVs are best served by a computationally efficient, vision-based approach to autonomous flight. In particular, in the DelFly project we have adopted a purposive vision approach to autonomous flight, in which only the information necessary for the robot’s task is extracted. The main goal of the approach is to allow the DelFly to autonomously explore unknown indoor environments, to which end we complement optical flow with other, appearance based, visual inputs.

This chapter deals with monocular obstacle detection. The visual cue of optic flow is used to determine the time-to-impact to obstacles in the environment. Since the flapping wing motion hampers the determination of optic flow, a complementary, “appearance variation cue” is studied. Combining these visual cues significantly improves detection results.

We introduce a novel setup of self-supervised learning (SSL), in which optical flow provides the supervised outputs. Optical flow requires significant movement for obstacle detection. The main advantage of the introduced method is that after learning, a robot can detect obstacles without moving - reducing the risk of collisions in narrow spaces. We investigate this novel setup of SSL in the context of a Micro Air Vehicle (MAV) that needs to select a suitable landing place. Initially, when the MAV flies over a potential landing area, the optical flow processing estimates a ‘surface roughness’ measure, capturing whether there are obstacles sticking out of the landing surface. This measure allows the MAV to select a safe landing place and then land with other optical flow measures such as the divergence. During flight, SSL takes place. For each image a texton distribution is extracted (capturing the visual appearance of the landing surface in sight), and mapped to the current roughness value by a linear regression function. We first demonstrate this principle to work with offline tests involving images captured on board an MAV, and then demonstrate the principle in flight. The experiments show that the MAV can land safely on the basis of optical flow. After learning it can also successfully select safe landing spots in hover. It is even shown that the appearance learning allows the pixel-wise segmentation of obstacles.

The autonomous capabilities of light-weight Flapping Wing Micro Air Vehicles (FWMAVs) have much to gain from onboard state estimation and attitude control. In this article, we present the first FWMAV with robust onboard state estimation and attitude control. The tailed FWMAV DelFly II was used, with the main goal to achieve active stabilization in the (passively unstable) hover condition. The attitude is estimated using an Inertial Measurement Unit with a gyroscope, accelerometer and magnetometer and the altitude is estimated using a barometer. A major challenge lies in the disturbance of the accelerometer measurements by the flapping motion of the wings. We propose a mechanical damping mechanism and flap-cycle based filtering to resolve this issue. The pitch estimates have a mean error of 1.5° with respect to the ground-truth measurement from a motion capture system. Using the onboard pitch estimate we can control the attitude of the FWMAV in the forward flight regime with a 30% lower standard deviation than in a trimmed flight. With a different set of gains, the FWMAV is able to perform a hovering flight - showing that a tailed FWMAV has enough control authority for this task. In a fully autonomous hover experiment, the DelFly II stays within a sphere of 0.75 m radius.

A new method is presented to decouple the parameters of the incident e(-) beam hitting the target of the linear accelerator, which consists essentially in optimizing the agreement between measurements and calculations when the difference filter, which is an additional filter inserted in the linac head to obtain uniform lateral dose-profile curves for the high energy photon beam, and flattening filter are removed from the beam path. This leads to lateral dose-profile curves, which depend only on the mean energy of the incident electron beam, since the effect of the radial intensity distribution of the incident e- beam is negligible when both filters are absent. The location of the primary collimator and the thickness and density of the target are not considered as adjustable parameters, since a satisfactory working Monte Carlo model is obtained for the low energy photon beam (6 MV) of the linac using the same target and primary collimator. This method was applied to conclude that the mean energy of the incident e- beam for the high energy photon beam (18 MV) of our Elekta SLi Plus linac is equal to 14.9 MeV. After optimizing the mean energy, the modelling of the filters, in accordance with the information provided by the manufacturer, can be verified by positioning only one filter in the linac head while the other is removed. It is also demonstrated that the parameter setting for Bremsstrahlung angular sampling in BEAMnrc ('Simple' using the leading term of the Koch and Motz equation or 'KM' using the full equation) leads to different dose-profile curves for the same incident electron energy for the studied 18 MV beam. It is therefore important to perform the calculations in 'KM' mode. Note that both filters are not physically removed from the linac head. All filters remain present in the linac head and are only rotated out of the beam. This makes the described method applicable for practical usage since no recommissioning process is required.

The purpose of this study was to assess the absorbed dose in and around lung tissue by performing radiochromic film measurements, Monte Carlo simulations and calculations with superposition convolution algorithms. We considered a layered polystyrene phantom of 12 x 12 x 12 cm3 containing a central cavity of 6 x 6 x 6 cm3 filled with Gammex RMI lung-equivalent material. Two field configurations were investigated, a small 1 x 10 cm2 field and a larger 10 x 10 cm2 field. First, we performed Monte Carlo simulations to investigate the influence of radiochromic film itself on the measured dose distribution when the film intersects a lung-equivalent region and is oriented parallel to the central beam axis. To that end, the film and the lung-equivalent materials were modelled in detail, taking into account their specific composition. Next, measurements were performed with the film oriented both parallel and perpendicular to the central beam axis to verify the results of our Monte Carlo simulations. Finally, we digitized the phantom in two commercially available treatment planning systems, Helax-TMS version 6.1A and Pinnacle version 6.2b, and calculated the absorbed dose in the phantom with their incorporated superposition convolution algorithms to compare with the Monte Carlo simulations. Comparing Monte Carlo simulations with measurements reveals that radiochromic film is a reliable dosimeter in and around lung-equivalent regions when the film is positioned perpendicular to the central beam axis. Radiochromic film is also able to predict the absorbed dose accurately when the film is positioned parallel to the central beam axis through the lung-equivalent region. However, attention must be paid when the film is not positioned along the central beam axis, in which case the film gradually attenuates the beam and decreases the dose measured behind the cavity. This underdosage disappears by offsetting the film a few centimetres. We find deviations of about 3.6% between Monte Carlo and the superposition convolution algorithm of Pinnacle behind the lung region, for both field configurations. Pinnacle is quite accurate in the lung region. Deviations up to 5.6% for the small field are found in the lung region between Monte Carlo and the superposition convolution algorithm of Helax-TMS. Behind the lung region, Helax-TMS is in better agreement with Monte Carlo. Radiochromic film measurements or Monte Carlo simulations are reliable methods to establish the dose in and around lung tissue.

Two Monte Carlo dose engines for radiotherapy treatment planning, namely a beta release of Peregrine and MCDE (Monte Carlo dose engine), were compared with Helax-TMS (collapsed cone superposition convolution) for a head and neck patient for the Elekta SLi plus linear accelerator. Deviations between the beta release of Peregrine and MCDE up to 10% were obtained in the dose volume histogram of the optical chiasm. It was illustrated that the differences are not caused by the particle transport in the patient, but by the modelling of the Elekta SLi plus accelerator head and more specifically the multileaf collimator (MLC). In MCDE two MLC modules (MLCQ and MLCE) were introduced to study the influence of the tongue-and-groove geometry, leaf bank tilt and leakage on the actual dose volume histograms. Differences in integral dose in the optical chiasm up to 3% between the two modules have been obtained. For single small offset beams though the FWHM of lateral profiles obtained with MLCE can differ by more than 1.5 mm from profiles obtained with MLCQ. Therefore, and because the recent version of MLCE is as fast as MLCQ, we advise to use MLCE for modelling the Elekta MLC. Nevertheless there still remains a large difference (up to 10%) between Peregrine and MCDE. By studying small offset beams we have shown that the profiles obtained with Peregrine are shifted, too wide and too flat compared with MCDE and phantom measurements. The overestimated integral doses for small beam segments explain the deviations observed in the dose volume histograms. The Helax-TMS results are in better agreement with MCDE, although deviations exceeding 5% have been observed in the optical chiasm. Monte Carlo dose deviations of more than 10% as found with Peregrine are unacceptable as an influence on the clinical outcome is possible and as the purpose of Monte Carlo treatment planning is to obtain an accuracy of 2%. We would like to emphasize that only the Elekta MLC has been tested in this work, so it is certainly possible that alpha releases of Peregrine provide more accurate results for other accelerators.