Reinforcement learning is considered to be a strong AI paradigm which can be used to teach machines through interaction with the environment and learning from their mistakes. Despite its perceived utility, it has not yet been successfully applied in automotive applications. Motivated by the successful demonstrations of learning of Atari games and Go by Google DeepMind, we propose a framework for autonomous driving using deep reinforcement learning. This is of particular relevance as it is difficult to pose autonomous driving as a supervised learning problem due to strong interactions with the environment including other vehicles, pedestrians and roadworks. As it is a relatively new area of research for autonomous driving, we provide a short overview of deep reinforcement learning and then describe our proposed framework. It incorporates Recurrent Neural Networks for information integration, enabling the car to handle partially observable scenarios. It also integrates the recent work on attention models to focus on relevant information, thereby reducing the computational complexity for deployment on embedded hardware. The framework was tested in an open source 3D car racing simulator called TORCS. Our simulation results demonstrate learning of autonomous maneuvering in a scenario of complex road curvatures and simple interaction of other vehicles.

Autonomous urban driving has a been rising problem since decades because of interactions with very complex environment. The traditional modular pipeline, rule-based algorithms, has not presented yet an efficient model to rely on, because it can not cover the very large possible scenarios space. Machine Learning techniques like supervised learning or imitation learning and Reinforcement learning have initial promising results with better performance. We propose an end-to-end Deep Conditional Imitation Learning model for autonomous driving inspired by both of Intel and Nvidia. The feature extraction part for Intel is replaced with Nvidia's model. Our proposed model outperforms Intel's architecture performance on CARLA Simulator, and overgeneralizes on various towns with different weather conditions.

Fully Autonomous Driving is considered as one of the difficult problems faced the automotive applications. It is forbidden due to the presence of some restricted Laws that prevent cars from being autonomous for the fear of accidents occurrence. However, researchers try to reach autonomous driving as a new area for research for the aim of having a strong push against these restricted Laws. Crash Avoidance functionality is one of the most important features in Self-Driving Cars that is partially integrated recently. We propose an end-to-end Crash Avoidance Deep IoT solution which is decomposed into two main parts: a) Detection Deep Neural Network which aims to detect accident occurrence in front of the ego-vehicle, and b) Accident Information Spreading IoT which is responsible for informing upcoming vehicles that there is an accident, then these vehicles will be able to take the reasonable actions either changing their routes, or changing their lanes avoiding crash. Due to the lack of Crash benchmarks, we build our own benchmark, depending only on a front camera, using ROS-Gazebo Simulation environment covering various crashes situations. In General, our proposed idea is the first solution that merges Deep Learning with IoT in automotive applications.

Currently, deep learning and IoT collaboration is heavily invading automotive applications especially in autonomous driving throughout successful assistance functionalities. Crash avoidance, path planning, and automatic emergency braking are essential functionalities for autonomous driving. Trigger-action-based IoT platforms are widely used due to its simplicity and ability of doing receptive tasks accurately. In this work, we propose SDC-Net system: an end-to-end deep learning IoT hybrid system in which a multitask neural network is trained based on different input representations from a camera-cocoon setup installed in CARLA simulator. We build our benchmark dataset covering different scenarios and corner cases that the vehicle may expose in order to navigate safely and robustly while testing. The proposed system aims to output relevant control actions for crash avoidance, path planning and automatic emergency braking. Multitask learning with a bird’s eye view input representation outperforms the nearest representation in precision, recall, f1-score, accuracy, and average MSE by more than 11.62%, 9.43%, 10.53%, 6%, and 25.84%, respectively.

Digital Automation Platforms are considered as a new category of programming without writing codes which increases the developer's access to many online services and IoT platforms through APIs. The proliferation of IoT, the connectivity of most of the devices to the internet and the growth of cloud-based services introduced new service-based automation platforms. Digital Platforms are a trigger-action programming that aim to make a set of workflows integrating different devices, provide useful functionality without having technical knowledge using graphical interface, automate human repetitive tasks, and offer certain types of decision-making capabilities. Many applications are integrated with these platforms, giving the user a high degree of freedom to create his own service via choosing both trigger and action. In this work, a comparative study is presented in order to differentiate between five automation platforms: Zapier, IFTTT, Integromat, Microsoft Power Automate and Parabola using various comparison metrics: accessibility, usability, time responsiveness with pricing model, integrability, and Security risks.

Few prior works study self-driving cars by deep learning with IoT collaboration. SDC-Net, which is an end-to-end multitask self-driving car camera cocoon IoT-based system, is one of the research areas that tackles this direction. However, by design, SDC-Net is not able to identify the accident locations; it only classifies whether a scene is a crash scene or not. In this work, we introduce an enhanced design for the SDC-Net system by (1) replacing the classification network with a detection one, (2) adapting our benchmark dataset labels built on the CARLA simulator to include the vehicles’ bounding boxes while keeping the same training, validation, and testing samples, and (3) modifying the shared information via IoT to include the accident location. We keep the same path planning and automatic emergency braking network, the digital automation platform, and the input representations to formulate the comparative study. The SDC-Net++ system is proposed to (1) output the relevant control actions, especially in case of accidents: accelerate, decelerate, maneuver, and brake, and (2) share the most critical information to the connected vehicles via IoT, especially the accident locations. A comparative study is also conducted between SDC-Net and SDC-Net++ with the same input representations: front camera only, panorama and bird’s eye views, and with single-task networks, crash avoidance only, and multitask networks. The multitask network with a BEV input representation outperforms the nearest representation in precision, recall, f1-score, and accuracy by more than 15.134%, 12.046%, 13.593%, and 5%, respectively. The SDC-Net++ multitask network with BEV outperforms SDC-Net multitask with BEV in precision, recall, f1-score, accuracy, and average MSE by more than 2.201%, 2.8%, 2.505%, 2%, and 18.677%, respectively.