Hung Manh La

Md Tamjid Hossain, Hung Manh La, Shahriar Badsha et al.

Although experience sharing (ES) accelerates multiagent reinforcement learning (MARL) in an advisor-advisee framework, attempts to apply ES to decentralized multiagent systems have so far relied on trusted environments and overlooked the possibility of adversarial manipulation and inference. Nevertheless, in a real-world setting, some Byzantine attackers, disguised as advisors, may provide false advice to the advisee and catastrophically degrade the overall learning performance. Also, an inference attacker, disguised as an advisee, may conduct several queries to infer the advisors' private information and make the entire ES process questionable in terms of privacy leakage. To address and tackle these issues, we propose a novel MARL framework (BRNES) that heuristically selects a dynamic neighbor zone for each advisee at each learning step and adopts a weighted experience aggregation technique to reduce Byzantine attack impact. Furthermore, to keep the agent's private information safe from adversarial inference attacks, we leverage the local differential privacy (LDP)-induced noise during the ES process. Our experiments show that our framework outperforms the state-of-the-art in terms of the steps to goal, obtained reward, and time to goal metrics. Particularly, our evaluation shows that the proposed framework is 8.32x faster than the current non-private frameworks and 1.41x faster than the private frameworks in an adversarial setting.

Adarsh Sehgal, Muskan Sehgal, Hung Manh La

Actor learning and critic learning are two components of the outstanding and mostly used Deep Deterministic Policy Gradient (DDPG) reinforcement learning method. Since actor and critic learning plays a significant role in the overall robot's learning, the performance of the DDPG approach is relatively sensitive and unstable as a result. We propose a multi-actor-critic DDPG for reliable actor-critic learning to further enhance the performance and stability of DDPG. This multi-actor-critic DDPG is then integrated with Hindsight Experience Replay (HER) to form our new deep learning framework called AACHER. AACHER uses the average value of multiple actors or critics to substitute the single actor or critic in DDPG to increase resistance in the case when one actor or critic performs poorly. Numerous independent actors and critics can also gain knowledge from the environment more broadly. We implemented our proposed AACHER on goal-based environments: AuboReach, FetchReach-v1, FetchPush-v1, FetchSlide-v1, and FetchPickAndPlace-v1. For our experiments, we used various instances of actor/critic combinations, among which A10C10 and A20C20 were the best-performing combinations. Overall results show that AACHER outperforms the traditional algorithm (DDPG+HER) in all of the actor/critic number combinations that are used for evaluation. When used on FetchPickAndPlace-v1, the performance boost for A20C20 is as high as roughly 3.8 times the success rate in DDPG+HER.

Adarsh Sehgal, Muskan Sehgal, Hung Manh La et al.

Accurate breast cancer diagnosis through mammography has the potential to save millions of lives around the world. Deep learning (DL) methods have shown to be very effective for mass detection in mammograms. Additional improvements of current DL models will further improve the effectiveness of these methods. A critical issue in this context is how to pick the right hyperparameters for DL models. In this paper, we present GA-E2E, a new approach for tuning the hyperparameters of DL models for brest cancer detection using Genetic Algorithms (GAs). Our findings reveal that differences in parameter values can considerably alter the area under the curve (AUC), which is used to determine a classifier's performance.

Ashutosh Singandhupe, Sanket Lokhande, Hung Manh La

Point cloud registration is a fundamental problem in computer vision and robotics, involving the alignment of 3D point sets captured from varying viewpoints using depth sensors such as LiDAR or structured light. In modern robotic systems, especially those focused on mapping, it is essential to merge multiple views of the same environment accurately. However, state-of-the-art registration techniques often struggle when large rotational differences exist between point sets or when the data is significantly corrupted by sensor noise. These challenges can lead to misalignments and, consequently, to inaccurate or distorted 3D reconstructions. In this work, we address both these limitations by proposing a robust modification to the classic Iterative Closest Point (ICP) algorithm. Our method, termed Exponential Similarity Matrix ICP (ESM-ICP), integrates a Gaussian-inspired exponential weighting scheme to construct a similarity matrix that dynamically adapts across iterations. This matrix facilitates improved estimation of both rotational and translational components during alignment. We demonstrate the robustness of ESM-ICP in two challenging scenarios: (i) large rotational discrepancies between the source and target point clouds, and (ii) data corrupted by non-Gaussian noise. Our results show that ESM-ICP outperforms traditional geometric registration techniques as well as several recent learning-based methods. To encourage reproducibility and community engagement, our full implementation is made publicly available on GitHub. https://github.com/aralab-unr/ESM_ICP

Adarsh Sehgal, Nicholas Ward, Hung Manh La et al.

Agents can base decisions made using reinforcement learning (RL) on a reward function. The selection of values for the learning algorithm parameters can, nevertheless, have a substantial impact on the overall learning process. In order to discover values for the learning parameters that are close to optimal, we extended our previously proposed genetic algorithm-based Deep Deterministic Policy Gradient and Hindsight Experience Replay approach (referred to as GA+DDPG+HER) in this study. On the robotic manipulation tasks of FetchReach, FetchSlide, FetchPush, FetchPick&Place, and DoorOpening, we applied the GA+DDPG+HER methodology. Our technique GA+DDPG+HER was also used in the AuboReach environment with a few adjustments. Our experimental analysis demonstrates that our method produces performance that is noticeably better and occurs faster than the original algorithm. We also offer proof that GA+DDPG+HER beat the current approaches. The final results support our assertion and offer sufficient proof that automating the parameter tuning procedure is crucial and does cut down learning time by as much as 57%.

Son Thanh Nguyen, Hai Nguyen, Son Tien Bui et al.

This paper presents a novel design of a multi-directional bicycle robot, which targets inspecting general ferromagnetic structures including complex-shaped structures. The locomotion concept is based on arranging two magnetic wheels in a bicycle-like configuration with two independent steering actuators. This configuration allows the robot to possess multi-directional mobility. An additional free joint helps the robot naturally adapt to non-flat and complex surfaces of steel structures. The robot has the biggest advantage to be mechanically simple with high mobility. Besides, the robot is equipped with sensing tools for structure health monitoring. We demonstrate the deployment of our robot to perform steel rust detection on steel bridges. The final inspection results are visualized as 3D models of the bridges together with marked locations of detected rusty areas.

Hoang-Dung Bui, Hung Manh La

Autonomous navigation of steel bridge inspection robots is essential for proper maintenance. The majority of existing robotic solutions for steel bridge inspection requires human intervention to assist in the control and navigation. In this paper, a control and navigation framework has been proposed for the steel bridge inspection robot developed by the Advanced Robotics and Automation (ARA)to facilitate autonomous real-time navigation and minimize human intervention. The ARA robot is designed to work in two modes: mobile and inch-worm. The robot uses mobile mode when moving on a plane surface and inch-worm mode when jumping from one surface to the other. To allow the ARA robot to switch between mobile and inch-worm modes, a switching controller is developed with 3D point cloud data based. The surface detection algorithm is proposed to allow the robot to check the availability of steel surfaces (plane, area, and height) to determine the transformation from mobile mode to inch-worm one. To have the robot safely navigate and visit all steel members of the bridge, four algorithms are developed to process the data from a depth camera, segment it into clusters, estimate the boundaries, construct a graph representing the structure, generate the shortest inspection path with any starting and ending points, and determine available robot configuration for path planning. Experiments on steel bridge structures setup highlight the effective performance of the algorithms, and the potential to apply to the ARA robot to run on real bridge structures.

Hoang-Dung Bui, Hai Nguyen, Hung Manh La et al.

Robot manipulation and grasping mechanisms have received considerable attention in the recent past, leading to the development of wide range of industrial applications. This paper proposes the development of an autonomous robotic grasping system for object sorting application. RGB-D data is used by the robot for performing object detection, pose estimation, trajectory generation, and object sorting tasks. The proposed approach can also handle grasping certain objects chosen by users. Trained convolutional neural networks are used to perform object detection and determine the corresponding point cloud cluster of the object to be grasped. From the selected point cloud data, a grasp generator algorithm outputs potential grasps. A grasp filter then scores these potential grasps, and the highest-scored grasp is chosen to execute on a real robot. A motion planner generates collision-free trajectories to execute the chosen grasp. The experiments on AUBO robotic manipulator show the potentials of the proposed approach in the context of autonomous object sorting with robust and fast sorting performance.

Adarsh Sehgal, Ashutosh Singandhupe, Hung Manh La et al.

Lidar-Monocular Visual Odometry (LIMO), a odometry estimation algorithm, combines camera and LIght Detection And Ranging sensor (LIDAR) for visual localization by tracking camera features as well as features from LIDAR measurements, and it estimates the motion using Bundle Adjustment based on robust key frames. For rejecting the outliers, LIMO uses semantic labelling and weights of the vegetation landmarks. A drawback of LIMO as well as many other odometry estimation algorithms is that it has many parameters that need to be manually adjusted according to the dynamic changes in the environment in order to decrease the translational errors. In this paper, we present and argue the use of Genetic Algorithm to optimize parameters with reference to LIMO and maximize LIMO's localization and motion estimation performance. We evaluate our approach on the well known KITTI odometry dataset and show that the genetic algorithm helps LIMO to reduce translation error in different datasets.

Adarsh Sehgal, Hung Manh La, Sushil J. Louis et al.

Reinforcement learning (RL) enables agents to take decision based on a reward function. However, in the process of learning, the choice of values for learning algorithm parameters can significantly impact the overall learning process. In this paper, we use a genetic algorithm (GA) to find the values of parameters used in Deep Deterministic Policy Gradient (DDPG) combined with Hindsight Experience Replay (HER), to help speed up the learning agent. We used this method on fetch-reach, slide, push, pick and place, and door opening in robotic manipulation tasks. Our experimental evaluation shows that our method leads to better performance, faster than the original algorithm.

Chuong Le, Huy Xuan Pham, Hung Manh La

There is a lot of waste in an industrial environment that could cause harmful effects to both the products and the workers resulting in product defects, itchy eyes or chronic obstructive pulmonary disease, etc. While automative cleaning robots could be used, the environment is often too big for one robot to clean alone in addition to the fact that it does not have adequate stored dirt capacity. We present a multi-robotic dirt cleaning system algorithm for multiple automatic iRobot Creates teaming to efficiently clean an environment. Moreover, since some spaces in the environment are clean while others are dirty, our multi-robotic system possesses a path planning algorithm to allow the robot team to clean efficiently by spending more time on the area with higher dirt level. Overall, our multi-robotic system outperforms the single robot system in time efficiency while having almost the same total battery usage and cleaning efficiency result.

Mehdi Rahimi, Spencer Gibb, Yantao Shen et al.

In this paper, a comparison of reinforcement learning algorithms and their performance on a robot box pushing task is provided. The robot box pushing problem is structured as both a single-agent problem and also a multi-agent problem. A Q-learning algorithm is applied to the single-agent box pushing problem, and three different Q-learning algorithms are applied to the multi-agent box pushing problem. Both sets of algorithms are applied on a dynamic environment that is comprised of static objects, a static goal location, a dynamic box location, and dynamic agent positions. A simulation environment is developed to test the four algorithms, and their performance is compared through graphical explanations of test results. The comparison shows that the newly applied reinforcement algorithm out-performs the previously applied algorithms on the robot box pushing problem in a dynamic environment.

Hai Nguyen, Hung Manh La, Matthew Deans

Supervised learning, more specifically Convolutional Neural Networks (CNN), has surpassed human ability in some visual recognition tasks such as detection of traffic signs, faces and handwritten numbers. On the other hand, even state-of-the-art reinforcement learning (RL) methods have difficulties in environments with sparse and binary rewards. They requires manually shaping reward functions, which might be challenging to come up with. These tasks, however, are trivial to human. One of the reasons that human are better learners in these tasks is that we are embedded with much prior knowledge of the world. These knowledge might be either embedded in our genes or learned from imitation - a type of supervised learning. For that reason, the best way to narrow the gap between machine and human learning ability should be to mimic how we learn so well in various tasks by a combination of RL and supervised learning. Our method, which integrates Deep Deterministic Policy Gradients and Hindsight Experience Replay (RL method specifically dealing with sparse rewards) with an experience ranking CNN, provides a significant speedup over the learning curve on simulated robotics tasks. Experience ranking allows high-reward transitions to be replayed more frequently, and therefore help learn more efficiently. Our proposed approach can also speed up learning in any other tasks that provide additional information for experience ranking.

Son Thanh Nguyen, Hung Manh La

Motivated by a high demand for automated inspection of civil infrastructure, this work presents a new design and development of a tank-like robot for structural health monitoring. Unlike most existing magnetic wheeled mobile robot designs, which may be suitable for climbing on flat steel surface, our proposed tank-like robot design uses reciprocating mechanism and roller-chains to make it capable of climbing on different structural shapes (e.g., cylinder, cube) with coated or non-coated steel surfaces. The proposed robot is able to transition from one surface to the other (e.g., from flat surface to curving surface). Taking into account of several strict considerations (including tight dimension, efficient adhesion and climbing flexibility) to adapt with various shapes of steel structures, a prototype tank-like robot incorporating multiple sensors (hall-effects, sonars, inertial measurement unit and camera), has been developed. Rigorous analysis of robot kinematics, adhesion force, sliding failure and turn-over failure has been conducted to demonstrate the stability of the proposed design. Mechanical and magnetic force analysis together with sliding/turn-over failure investigation can serve as an useful framework for designing various steel climbing robots in the future. Experimental results and field deployments confirm the adhesion and climbing capability of the developed robot.

Huy Xuan Pham, Hung Manh La, David Feil-Seifer et al.

This paper proposes a distributed Multi-Agent Reinforcement Learning (MARL) algorithm for a team of Unmanned Aerial Vehicles (UAVs). The proposed MARL algorithm allows UAVs to learn cooperatively to provide a full coverage of an unknown field of interest while minimizing the overlapping sections among their field of views. Two challenges in MARL for such a system are discussed in the paper: firstly, the complex dynamic of the joint-actions of the UAV team, that will be solved using game-theoretic correlated equilibrium, and secondly, the challenge in huge dimensional state space representation will be tackled with efficient function approximation techniques. We also provide our experimental results in detail with both simulation and physical implementation to show that the UAV team can successfully learn to accomplish the task.

Huy Xuan Pham, Hung Manh La, David Feil-Seifer et al.

Wild-land fire fighting is a hazardous job. A key task for firefighters is to observe the "fire front" to chart the progress of the fire and areas that will likely spread next. Lack of information of the fire front causes many accidents. Using Unmanned Aerial Vehicles (UAVs) to cover wildfire is promising because it can replace humans in hazardous fire tracking and significantly reduce operation costs. In this paper we propose a distributed control framework designed for a team of UAVs that can closely monitor a wildfire in open space, and precisely track its development. The UAV team, designed for flexible deployment, can effectively avoid in-flight collisions and cooperate well with neighbors. They can maintain a certain height level to the ground for safe flight above fire. Experimental results are conducted to demonstrate the capabilities of the UAV team in covering a spreading wildfire.

Tuan Le, Spencer Gibb, Hung Manh La et al.

Bridge condition assessment is important to maintain the quality of highway roads for public transport. Bridge deterioration with time is inevitable due to aging material, environmental wear and in some cases, inadequate maintenance. Non-destructive evaluation (NDE) methods are preferred for condition assessment for bridges, concrete buildings, and other civil structures. Some examples of NDE methods are ground penetrating radar (GPR), acoustic emission, and electrical resistivity (ER). NDE methods provide the ability to inspect a structure without causing any damage to the structure in the process. In addition, NDE methods typically cost less than other methods, since they do not require inspection sites to be evacuated prior to inspection, which greatly reduces the cost of safety related issues during the inspection process. In this paper, an autonomous robotic system equipped with three different NDE sensors is presented. The system employs GPR, ER, and a camera for data collection. The system is capable of performing real-time, cost-effective bridge deck inspection, and is comprised of a mechanical robot design and machine learning and pattern recognition methods for automated steel rebar picking to provide realtime condition maps of the corrosive deck environments.

Devin Connell, Hung Manh La

It is necessary for a mobile robot to be able to efficiently plan a path from its starting, or current, location to a desired goal location. This is a trivial task when the environment is static. However, the operational environment of the robot is rarely static, and it often has many moving obstacles. The robot may encounter one, or many, of these unknown and unpredictable moving obstacles. The robot will need to decide how to proceed when one of these obstacles is obstructing it's path. A method of dynamic replanning using RRT* is presented. The robot will modify it's current plan when an unknown random moving obstacle obstructs the path. Various experimental results show the effectiveness of the proposed method.

Ashutosh Singandhupe, Hung Manh La, David Feil-Seifer et al.

Unmanned aerial vehicles (UAVs) have gained much attention in recent years for both commercial and military applications. The progress in this field has gained much popularity and the research has encompassed various fields of scientific domain. Cyber securing a UAV communication has been one of the active research field since the attack on Predator UAV video stream hijacking in 2009. Since UAVs rely heavily on on-board autopilot to function, it is important to develop an autopilot system that is robust to possible cyber attacks. In this work, we present a biometric system to encrypt the UAV communication by generating a key which is derived from Beta component of the EEG signal of a user. We have developed a safety mechanism that would be activated in case the communication of the UAV from the ground control station gets attacked. This system has been validated on a commercial UAV under malicious attack conditions during which we implement a procedure where the UAV return safely to a "home" position.