Background: Nowadays, logistics service providers face several challenges which create an urge to rethink their strategy to improve their position within the market,decrease their costs and their environmental impact. At the same time theintroduction of autonomous driving potentially has an impact on logistics.Self-driving trucks can help logistics companies to tackle these challenges.However, the implementation of this technology could fundamentally alterthe competitive landscape. Hence, certain competitive advantages currentlyheld by logistics firms might lose their relevance in the future and need tobe adapted to maintain a strong market position. Purpose: The purpose of this study is to explore the perception of self-driving trucks within logistics and the impact on competitive advantages of logistics service providers. Thereby, this thesis will look at experts from Germany and Sweden and their opinion on future implications of self-driving trucks. Method: An inductive research approach is used to explore the topic. A multi-method research strategy is applied to gather data through qualitative semi-structured interviews with 17 participants. These were divided into five different case groups. To interpret the data a thematic analysis approach was chosen. Conclusion: The main contribution is a model representing the impact of autonomous driving on competitive advantages and the implications for the logistics industry. Findings are based on the perception of experts about autonomous driving, current resources and capabilities.

This work investigates the problem of transfer from simulation to the real world in the context of autonomous navigation. To this end, we first present a photo-realistic training and evaluation simulator (Sim4CV)* which enables several applications across various fields of computer vision. Built on top of the Unreal Engine, the simulator features cars and unmanned aerial vehicles (UAVs) with a realistic physics simulation and diverse urban and suburban 3D environments. We demonstrate the versatility of the simulator with two case studies: autonomous UAV-based tracking of moving objects and autonomous driving using supervised learning. Using the insights gained from aerial object tracking, we find that current object trackers are either too slow or inaccurate for online tracking from an UAV. In addition, we find that in particular background clutter, fast motion and occlusion are preventing fast trackers such as correlation filter (CF) trackers to perform better. To address this issue we propose a novel and general framework that can be applied to CF trackers in order incorporate context. As a result the learned filter is more robust to drift due to the aforementioned tracking challenges. We show that our framework can improve several CF trackers by a large margin while maintaining a very high frame rate. For the application of autonomous driving, we train a driving policy that drives very well in simulation. However, while our simulator is photo-realistic there still exists a virtual-reality gap. We show how this gap can be reduced via modularity and abstraction in the driving policy. More specifically, we split the driving task into several modules namely perception, driving policy and control. This simplifies the transfer significantly and we show how a driving policy that was only trained in simulation can be transferred to a robotic vehicle in the physical world directly. Lastly, we investigate the application of UAV racing which has emerged as a modern sport recently. We propose a controller fusion network (CFN) which allows fusing multiple imperfect controllers; the result is a navigation policy that outperforms each one of them. Further, we embed this CFN into a modular network architecture similar to the one for driving, in order to decouple perception and control. We use our photo-realistic simulation environment to demonstrate how navigation policies can be transferred to different environment conditions by this network modularity.

No description available.

Thanks to the major advancements in hardware and computational power, sensor technology, and artificial intelligence, the race for fully autonomous driving systems is heating up. With a countless number of challenging conditions and driving scenarios, researchers are tackling the most challenging problems in driverless cars. One of the most critical components is the perception module, which enables an autonomous vehicle to "see" and "understand" its surrounding environment. Given that modern vehicles can have large number of sensors and available data streams, this thesis presents a deep learning-based framework that leverages multimodal data – i.e. sensor fusion, to perform the task of 3D object detection and localization. We provide an extensive review of the advancements of deep learning-based methods in computer vision, specifically in 2D and 3D object detection tasks. We also study the progress of the literature in both single-sensor and multi-sensor data fusion techniques. Furthermore, we present an in-depth explanation of our proposed approach that performs sensor fusion using input streams from LiDAR and Camera sensors, aiming to simultaneously perform 2D, 3D, and Bird’s Eye View detection. Our experiments highlight the importance of learnable data fusion mechanisms and multi-task learning, the impact of different CNN design decisions, speed-accuracy tradeoffs, and ways to deal with overfitting in multi-sensor data fusion frameworks.

An IMU is a sensor with many differing use cases, it makes use of an accelerometer, gyroscope and sometimes a magnetometer. One of the biggest problems with IMU sensors is the effect vibrations can have on their data. The reason for this study is to find a solution to this problem by filtering the data. The tests for this study were conducted in cooperation with Husqvarna using two of their automowers. The tests were made by running the automowers across different surfaces and recording the IMU data. To find filters for the IMU data a comprehensive literature survey was conducted to find suitable methods to filter out vibrations. The two filters selected for further testing were the complementary filter and the LMS filter. When the tests had been run all the data was added to data sheets where it could be analyzed and have the filters added to the data. From the gathered data the data spikes were clearly visible and were more than enough to trigger the mower's emergency stop and need to be manually reset. The vibrations were too irregular to filter using the LMS filter since it requires a known signal to filter against. Hence only the complementary filter was implemented fully. With the complementary filter these vibrations can be minimized and brought well below the level required to trigger an emergency stop. With a high filter weight constant such as 0.98, the margin of error from vibrations can be brought down to +- 1 degrees as the lowest and +- 4,6 degrees as highest depending on the surface and automower under testing. The main advantage with using the complementary filter is that it only requires one weight constant to adjust the filter intensity making it easy to use. The one disadvantage is that the higher the weight constant is the more delay there is on the data.

The master's thesis deals with the design and realization of embedded control system and software of the autonomous mobile robot DACEP. The research section focuses on the selection of sensory equipment. Moreover, the design of the embedded control system and the communication interface between this system and the master PC is described in detail, followed by the design of localization and navigation software that uses ROS framework. The section is written as instructive as possible for the development of robots of similar construction. Finally the development of a graphical interface for robot diagnostics and remote control is depicted.

<div><div><div><p>Towing capacity affects a vehicle’s towing ability and it is usually costly to buy or even rent a vehicle that can tow certain amount of weight. A widely swaying towing trailer is one of the main causes for accidents that involves towing trailers. This study propose an affordable automated disconnected towing system (ADTS) that does not require physical connection between leading vehicle and the trailer vehicle by only using a computer vision system. The ADTS contains two main parts: a leading vehicle which can perform lane detection and a trailer vehicle which can automatically follow the leading vehicle by detecting the license plate of the leading vehicle. The trailer vehicle can adjust its speed according to the distance from the leading vehicle.</p></div></div></div>

Background. Machine learning technology is used daily in many aspects of computers. Neural network is a machine learning technique. The importance of carsthat are self driving has increased in recent years and the research about it has alsoincreased. Some cars that are produced today already have an autopilot feature inthem. Objectives. In this thesis the objective is to find how the training duration of aneural network changes, when the track becomes harder and harder. Methods. To achieve this a small game and a neural network was developed inUnity. The neural network receive input from 5 sensors and gave 2 outputs, an angleand a velocity. The neural network will be trained in the game on 4 different tracks,each with one more obstacle than the last one. The neural network will generatetwenty cars that will drive simultaneously to the end of the track. If a car collidewith a wall, it gets destroyed. When all cars are destroyed a new generation is generated based on the 2 cars that got destroyed last. Results. The first track was finished in 8,3 seconds, the second track in 28,9 secondsand that is a 248% increase from the first track. The third track was finished in 44,5seconds and that is a 54% increase from the second track. Fourth track were finishedin 39,9 seconds and that was a 10,3% decrease in time. Conclusions. In conclusion the training duration increased from track 1 to track2. from track 2 to track 3 and 4 the increase is presumed to be because the trackis larger and because of the speed output from the neural network. The trainingduration does change, but the change corresponds with the change of the length ofthe track.

This degree project investigates the suitability of using Behavior Trees (BT) as an architecture for the behavioral layer in autonomous driving. BTs originate from video game development but have received attention in robotics research the past couple of years. This project also includes implementation of a simulated traffic environment using the Unity3D engine, where the use of BTs is evaluated and compared to an implementation using finite-state machines (FSM). After the initial implementation, the simulation along with the control architectures were extended with additional behaviors in four steps. The different versions were evaluated using software maintainability metrics (Cyclomatic complexity and Maintainability index) in order to extrapolate and reason about more complex implementations as would be required in a real autonomous vehicle. It is concluded that as the AI requirements scale and grow more complex, the BTs likely become substantially more maintainable than FSMs and hence may prove a viable alternative for autonomous driving.

Self-driving cars have recently become a challenging research topic, with the aim of making transportation safer and more efficient. Current advanced driving assistance systems (ADAS) allow cars to drive autonomously by following lane markings, identifying road signs and detecting pedestrians and other vehicles. In this thesis work we improve the robustness of autonomous cars by designing an on-road obstacle detection system. The proposed solution consists on the low-level fusion of radar and lidar through the occupancy grid framework. Two inference theories are implemented and evaluated: Bayesian probability theory and Dempster-Shafer theory of evidence. Obstacle detection is performed through image processing of the occupancy grid. Last, the Dempster-Shafer additional features are leveraged by proposing a sensor performance estimation module and performing advanced conflict management. The work has been carried out at Volvo Car Corporation, where real experiments on a test vehicle have been performed under different environmental conditions and types of objects. The system has been evaluated according to the quality of the resulting occupancy grids, detection rate as well as information content in terms of entropy. The results show a significant improvement of the detection rate over single-sensor approaches. Furthermore, the Dempster-Shafer implementation may slightly outperform the Bayesian one when there is conflicting information, although the high computational cost limits its practical application. Last, we demonstrate that the proposed solution is easily scalable to include additional sensors.