An agent that efficiently accumulates knowledge to develop increasingly sophisticated skills over a long lifetime could advance the frontier of artificial intelligence capabilities. The design of such agents, which remains a long-standing challenge of artificial intelligence, is addressed by the subject of continual learning. This monograph clarifies and formalizes concepts of continual learning, introducing a framework and set of tools to stimulate further research.

Spring-based actuators in legged locomotion provide energy-efficiency and improved performance, but increase the difficulty of controller design. While previous work has focused on extensive modeling and simulation to find optimal controllers for such systems, we propose to learn model-free controllers directly on the real robot. In our approach, gaits are first synthesized by central pattern generators (CPGs), whose parameters are optimized to quickly obtain an open-loop controller that achieves efficient locomotion. Then, to make this controller more robust and further improve the performance, we use reinforcement learning to close the loop, to learn corrective actions on top of the CPGs. We evaluate the proposed approach on the DLR elastic quadruped bert. Our results in learning trotting and pronking gaits show that exploitation of the spring actuator dynamics emerges naturally from optimizing for dynamic motions, yielding high-performing locomotion, particularly the fastest walking gait recorded on bert, despite being model-free. The whole process takes no more than 1.5 hours on the real robot and results in natural-looking gaits.

Scheduling on dataflow graphs (also known as computation graphs) is an NP-hard problem. The traditional exact methods are limited by runtime complexity, while reinforcement learning (RL) and heuristic-based approaches struggle with determinism and solution quality. This research aims to develop an innovative approach that employs machine learning (ML) for addressing combinatorial optimization problems, using scheduling as a case study. The goal is to provide guarantees in optimality and determinism while maintaining the runtime cost of heuristic methods. Specifically, we introduce a novel two-phase RL-to-ILP scheduling framework, which includes three steps: 1) RL solver acts as coarse-grain scheduler, 2) solution relaxation and 3) exact solving via ILP. Our framework demonstrates the same scheduling performance compared with using exact scheduling methods while achieving up to 128 $\times$ speed improvements. This was conducted on actual EdgeTPU platforms, utilizing ImageNet DNN computation graphs as input. Additionally, the framework offers improved on-chip inference runtime and acceleration compared to the commercially available EdgeTPU compiler.

Internet of Things (IoT) technologies have enabled numerous data-driven mobile applications and have the potential to significantly improve environmental monitoring and hazard warnings through the deployment of a network of IoT sensors. However, these IoT devices are often power-constrained and utilize wireless communication schemes with limited bandwidth. Such power constraints limit the amount of information each device can share across the network, while bandwidth limitations hinder sensors' coordination of their transmissions. In this work, we formulate the communication planning problem of IoT sensors that track the state of the environment. We seek to optimize sensors' decisions in collecting environmental data under stringent resource constraints. We propose a multi-agent reinforcement learning (MARL) method to find the optimal communication policies for each sensor that maximize the tracking accuracy subject to the power and bandwidth limitations. MARL learns and exploits the spatial-temporal correlation of the environmental data at each sensor's location to reduce the redundant reports from the sensors. Experiments on wildfire spread with LoRA wireless network simulators show that our MARL method can learn to balance the need to collect enough data to predict wildfire spread with unknown bandwidth limitations.

Machine learning has proven to be a useful tool for automated malware detection, but machine learning models have also been shown to be vulnerable to adversarial attacks. This article addresses the problem of generating adversarial malware samples, specifically malicious Windows Portable Executable files. We summarize and compare work that has focused on adversarial machine learning for malware detection. We use gradient-based, evolutionary algorithm-based, and reinforcement-based methods to generate adversarial samples, and then test the generated samples against selected antivirus products. We compare the selected methods in terms of accuracy and practical applicability. The results show that applying optimized modifications to previously detected malware can lead to incorrect classification of the file as benign. It is also known that generated malware samples can be successfully used against detection models other than those used to generate them and that using combinations of generators can create new samples that evade detection. Experiments show that the Gym-malware generator, which uses a reinforcement learning approach, has the greatest practical potential. This generator achieved an average sample generation time of 5.73 seconds and the highest average evasion rate of 44.11%. Using the Gym-malware generator in combination with itself improved the evasion rate to 58.35%.

In the realm of multi-agent reinforcement learning, intrinsic motivations have emerged as a pivotal tool for exploration. While the computation of many intrinsic rewards relies on estimating variational posteriors using neural network approximators, a notable challenge has surfaced due to the limited expressive capability of these neural statistics approximators. We pinpoint this challenge as the "revisitation" issue, where agents recurrently explore confined areas of the task space. To combat this, we propose a dynamic reward scaling approach. This method is crafted to stabilize the significant fluctuations in intrinsic rewards in previously explored areas and promote broader exploration, effectively curbing the revisitation phenomenon. Our experimental findings underscore the efficacy of our approach, showcasing enhanced performance in demanding environments like Google Research Football and StarCraft II micromanagement tasks, especially in sparse reward settings.

We consider a network of smart sensors for an edge computing application that sample a time-varying signal and send updates to a base station for remote global monitoring. Sensors are equipped with sensing and compute, and can either send raw data or process them on-board before transmission. Limited hardware resources at the edge generate a fundamental latency-accuracy trade-off: raw measurements are inaccurate but timely, whereas accurate processed updates are available after processing delay. Hence, one needs to decide when sensors should transmit raw measurements or rely on local processing to maximize network monitoring performance. To tackle this sensing design problem, we model an estimation-theoretic optimization framework that embeds both computation and communication latency, and propose a Reinforcement Learning-based approach that dynamically allocates computational resources at each sensor. Effectiveness of our proposed approach is validated through numerical experiments motivated by smart sensing for the Internet of Drones and self-driving vehicles. In particular, we show that, under constrained computation at the base station, monitoring performance can be further improved by an online sensor selection.

Overtaking on two-lane roads is a great challenge for autonomous vehicles, as oncoming traffic appearing on the opposite lane may require the vehicle to change its decision and abort the overtaking. Deep reinforcement learning (DRL) has shown promise for difficult decision problems such as this, but it requires massive number of data, especially if the action space is continuous. This paper proposes to incorporate guidance from an expert system into DRL to increase its sample efficiency in the autonomous overtaking setting. The guidance system developed in this study is composed of constrained iterative LQR and PID controllers. The novelty lies in the incorporation of a fading guidance function, which gradually decreases the effect of the expert system, allowing the agent to initially learn an appropriate action swiftly and then improve beyond the performance of the expert system. This approach thus combines the strengths of traditional control engineering with the flexibility of learning systems, expanding the capabilities of the autonomous system. The proposed methodology for autonomous vehicle overtaking does not depend on a particular DRL algorithm and three state-of-the-art algorithms are used as baselines for evaluation. Simulation results show that incorporating expert system guidance improves state-of-the-art DRL algorithms greatly in both sample efficiency and driving safety.

Deep Reinforcement Learning (Deep-RL) techniques for motion control have been continuously used to deal with decision-making problems for a wide variety of robots. Previous works showed that Deep-RL can be applied to perform mapless navigation, including the medium transition of Hybrid Unmanned Aerial Underwater Vehicles (HUAUVs). These are robots that can operate in both air and water media, with future potential for rescue tasks in robotics. This paper presents new approaches based on the state-of-the-art Double Critic Actor-Critic algorithms to address the navigation and medium transition problems for a HUAUV. We show that double-critic Deep-RL with Recurrent Neural Networks using range data and relative localization solely improves the navigation performance of HUAUVs. Our DoCRL approaches achieved better navigation and transitioning capability, outperforming previous approaches.

Deep reinforcement learning (DRL) techniques have become increasingly used in various fields for decision-making processes. However, a challenge that often arises is the trade-off between both the computational efficiency of the decision-making process and the ability of the learned agent to solve a particular task. This is particularly critical in real-time settings such as video games where the agent needs to take relevant decisions at a very high frequency, with a very limited inference time. In this work, we propose a generic offline learning approach where the computation cost of the input features is taken into account. We derive the Budgeted Decision Transformer as an extension of the Decision Transformer that incorporates cost constraints to limit its cost at inference. As a result, the model can dynamically choose the best input features at each timestep. We demonstrate the effectiveness of our method on several tasks, including D4RL benchmarks and complex 3D environments similar to those found in video games, and show that it can achieve similar performance while using significantly fewer computational resources compared to classical approaches.