Autonomous assembly is an essential capability for industrial and service robots, with Peg-in-Hole (PiH) insertion being one of the core tasks. However, PiH assembly in unknown environments is still challenging due to uncertainty in task parameters, such as the hole position and orientation, resulting from sensor noise. Although context-based meta reinforcement learning (RL) methods have been previously presented to adapt to unknown task parameters in PiH assembly tasks, the performance depends on a sample-inefficient procedure or human demonstrations. Thus, to enhance the applicability of meta RL in real-world PiH assembly tasks, we propose to train the agent to use information from the robot’s forward kinematics and an uncalibrated camera. Furthermore, we improve the applicability by efficiently adapting the meta-trained agent to use data from force/torque sensor. Finally, we propose an adaptation procedure for out-of-distribution tasks whose parameters are different from the training tasks. Experiments on simulated and real robots prove that our modifications enhance the sample efficiency during meta training, real-world adaptation performance, and generalization of the context-based meta RL agent in PiH assembly tasks compared to previous approaches.

Accurate floor localization in a multi-story environment is an important but challenging task. Among the current floor localization techniques, fingerprinting is the mainstream technology due to its accuracy in noisy environments. To achieve accurate floor localization in a building with many floors, we have to collect sufficient data on each floor, which needs significant storage and running time; preventing fingerprinting techniques from scaling to support large multi-story buildings, especially on a worldwide scale.In this paper, we propose a quantum algorithm for accurate multi-story localization. The proposed algorithm leverages quantum computing concepts to provide an exponential enhancement in both space and running time compared to the classical counterparts. In addition, it builds on an efficient binary-neuron implementation that can be implemented using fewer qubits compared to the typical non-binary neurons, allowing for easier deployment with near-term quantum devices. We implement the proposed algorithm on a real IBM quantum machine and evaluate it on three real indoor testbeds. Results confirm the exponential saving in both time and space for the proposed quantum algorithm, while keeping the same localization accuracy compared to the traditional classical techniques, and using half the number of qubits required for other quantum localization algorithms.

Effective access points (APs) selection is a crucial step in localization systems. It directly affects both localization accuracy and computational efficiency. Classical APs selection algorithms are usually computationally expensive, hindering the deployment of localization systems in a large worldwide scaleIn this paper, we introduce a quantum APs selection algorithm for large-scale localization systems. The proposed algorithm leverages quantum annealing to eliminate redundant and noisy APs. We explain how to formulate the APs selection problem as a quadratic unconstrained binary optimization (QUBO) problem, suitable for quantum annealing, and how to select the minimum number of APs that maintain the same overall localization system accuracy as the complete APs set. Based on this, we further propose a logarithmic-complexity algorithm to select the optimal number of APs.We implement our quantum algorithm on a real D-Wave Systems quantum machine and assess its performance in a real test environment for a floor localization problem. Our findings reveal that by selecting fewer than 14% of the available APs in the environment, our quantum algorithm achieves the same floor localization accuracy as utilizing the entire set of APs and a superior accuracy over utilizing the reduced dataset by classical APs selection counterparts. Moreover, the proposed quantum algorithm achieves more than an order of magnitude speedup over the corresponding classical APs selection algorithms, emphasizing the efficiency of the proposed quantum algorithm for large-scale localization systems.

The recent release of the third-generation partnership project, Release 17, calls for sub-meter cellular positioning accuracy with reduced latency in the calculation. To provide such high accuracy on a worldwide scale, leveraging the received signal strength (RSS) for positioning promises ubiquitous availability in the current and future equipment. RSS Fingerprint-based techniques have shown great potential for providing high accuracy in both indoor and outdoor environments. However, fingerprint-based positioning faces the challenge of providing a fast matching algorithm that can scale worldwide. In this paper, we propose a cosine similarity-based quantum algorithm for enabling fingerprint-based high accuracy and worldwide positioning that can be integrated with the next generation of 5G and 6G networks and beyond. By entangling the test RSS vector with the fingerprint RSS vectors, the proposed quantum algorithm has a complexity that is exponentially better than its classical version as well as the state-of-the-art quantum fingerprint positioning systems, both in the storage space and the running time. We implement the proposed quantum algorithm and evaluate it in a cellular testbed on a real IBM quantum machine. Results show the exponential saving in both time and space for the proposed quantum algorithm while keeping the same positioning accuracy compared to the traditional classical fingerprinting techniques and the state-of-the-art quantum algorithms.

Quantum computing is a fast-developing field that has the ability to tackle complex problems that conventional computers struggle with. Recently, quantum fingerprinting localization has been introduced, which allows for the implementation of large-scale location determination systems across the globe. In this paper, we introduce QRadar, a localization system that uses quantum similarity fingerprints based on the Euclidean similarity metric, which is an early and widely used measure. The computational complexity of QRadar is exponentially superior to classical systems. We explain how to generate the quantum fingerprint, how to encode the received signal strength (RSS) measurements as quantum particles, and we describe the quantum algorithm for computing the Euclidean similarity distance between the online RSS measurements and the fingerprint ones. Additionally, we investigate various sources of errors in quantum machines and how they affect the accuracy of QRadar. We then discuss how to choose an appropriate quantum machine to minimize localization errors.We installed QRadar on a real IBM Quantum Experience machine. The results we obtained from both the installation and simulations conducted on two real test sites confirm that QRadar can accurately determine the estimated location with a significant improvement in processing time when compared to conventional classical methods.

Fingerprinting is one of the mainstream technologies for localization. However, it needs significant storage overhead and running time, preventing it from scaling to support world-wide indoor/outdoor localization.Quantum computing has the potential to revolutionize computation by making some classically intractable problems solvable on quantum computers. In this paper, we propose a quantum fingerprint-based localization algorithm for enabling large-scale location tracking systems, envisioning future era of location tracking and spatial systems. Specifically, we propose a quantum algorithm that provides an exponential enhancement of both the space and running time complexity compared to the traditional classical systems. We give the details of how to build the quantum fingerprint, how to encode the received signal strength (RSS) measurements in quantum particles, and finally; present a quantum algorithm for calculating the cosine similarity between the online RSS measurements and the fingerprint ones.Results from deploying our algorithm in three real testbeds on IBM Quantum Experience machines confirm the ability of our quantum system to get the same accuracy as the classical one but with the potential exponential saving in both space and running time.

Abstract Brain tumor detection can make the difference between life and death. Recently, deep learning-based brain tumor detection techniques have gained attention due to their higher performance. However, obtaining the expected performance of such deep learning-based systems requires large amounts of classified images to train the deep models. Obtaining such data is usually boring, time-consuming, and can easily be exposed to human mistakes which hinder the utilization of such deep learning approaches. This paper introduces a novel framework for brain tumor detection and classification. The basic idea is to generate a large synthetic MRI images dataset that reflects the typical pattern of the brain MRI images from a small class-unbalanced collected dataset. The resulted dataset is then used for training a deep model for detection and classification. Specifically, we employ two types of deep models. The first model is a generative model to capture the distribution of the important features in a set of small class-unbalanced brain MRI images. Then by using this distribution, the generative model can synthesize any number of brain MRI images for each class. Hence, the system can automatically convert a small unbalanced dataset to a larger balanced one. The second model is the classifier that is trained using the large balanced dataset to detect brain tumors in MRI images. The proposed framework acquires an overall detection accuracy of 96.88% which highlights the promise of the proposed framework as an accurate low-overhead brain tumor detection system.

Abstract A new generalized framework for lung cancer detection and classification are introduced in this paper. Specifically, two types of deep models are presented. The first model is a generative model to capture the distribution of the important features in a set of small class-unbalanced collected CXR images. This generative model can be utilized to synthesize any number of CXR images for each class. For example, our generative model can generate images with tumors with different sizes and positions in the lung. Hence, the system can automatically convert the small unbalanced collected dataset to a larger balanced one. The second model is the ResNet50 that is trained using the large balanced dataset for cancer classification into benign and malignant. The proposed framework acquires 98.91% overall detection accuracy, 98.85% area under curve (AUC), 98.46% sensitivity, 97.72% precision, 97.89% F1 score. The classifier takes 1.2334 s on average to classify a single image using a machine with 13GB RAM.

Recently, deep learning-based localization systems have become one of the most promising techniques due to their accuracy in complex environments. However, these techniques require large amounts of data for training. Obtaining such data is usually a tedious and time-consuming process, which hinders their practical deployment. In this paper, we propose a data augmentation framework for deep learning-based localization systems. The basic idea is to use a conditional Generative Adversarial Network that is able to learn the complex structures in the original training data and then generate high-quality synthetic data that matches the original data distribution. Evaluation of the proposed data augmentation framework in a real testbed shows that our technique can increase the average localization accuracy by 22.2% compared to the case of not using data augmentation. This demonstrates the promise of the proposed framework for enhancing deep learning-based localization systems.

Quantum computing provides a new way for approaching problem solving, enabling efficient solutions for problems that are hard on classical computers. With researchers around the world showing quantum supremacy and the availability of cloud-based quantum computers, quantum computing is becoming a reality. In this paper, we explore the different directions of the use of quantum computing for location tracking and spatial systems. Specifically, we show an example for the expected gain of using quantum computing for localization by providing an efficient quantum algorithm for RF fingerprinting localization. The proposed quantum algorithm has a complexity that is exponentially better than its classical algorithm version, both in space and running time. We further discuss both software and hardware research challenges and opportunities that researchers can build on to explore this exciting new domain.

The ability to accurately evaluate the performance of location determination systems is crucial for many applications. Typically, the performance of such systems is obtained by comparing ground truth locations with estimated locations. However, these ground truth locations are usually obtained by clicking on a map or using other worldwide available technologies like GPS. This introduces ground truth errors that are due to the marking process, map distortions, or inherent GPS inaccuracy.In this paper, we present a theoretical framework for analyzing the effect of ground truth errors on the evaluation of localization systems. Based on that, we design two algorithms for computing the real algorithmic error from the validation error and marking/map ground truth errors, respectively. We further establish bounds on different performance metrics.Validation of our theoretical assumptions and analysis using real data collected in a typical environment shows the ability of our theoretical framework to correct the estimated error of a localization algorithm in the presence of ground truth errors. Specifically, our marking error algorithm matches the real error CDF within 4%, and our map error algorithm provides a more accurate estimate of the median/tail error by 150%/72% when the map is shifted by 6m.

We present DynamicSLAM: an indoor localization technique that eliminates the need for the daunting calibration step. DynamicSLAM is a novel Simultaneous Localization And Mapping (SLAM) framework that iteratively acquires the feature map of the environment while simultaneously localizing users relative to this map. Specifically, we employ the phone inertial sensors to keep track of the user's path. To compensate for the error accumulation due to the low-cost inertial sensors, DynamicSLAM leverages unique points in the environment (anchors) as observations to reduce the estimated location error. DynamicSLAM introduces the novel concept of mobile human anchors that are based on the encounters with other users in the environment, significantly increasing the number and ubiquity of anchors and boosting localization accuracy. We present different encounter models and show how they are incorporated in a unified probabilistic framework to reduce the ambiguity in the user location. Furthermore, we present a theoretical proof for system convergence and the human anchors ability to reset the accumulated error. Evaluation of DynamicSLAM using different Android phones shows that it can provide a localization accuracy with a median of 1.1m. This accuracy outperforms the state-of-the-art techniques by 55 percent, highlighting DynamicSLAM promise for ubiquitous indoor localization.

Recently, deep learning-based positioning systems have gained attention due to their higher performance relative to traditional methods. However, obtaining the expected performance of deep learning-based systems requires large amounts of data to train model. Obtaining this data is usually a tedious process which hinders the utilization of such deep learning approaches. In this paper, we introduce a number of techniques for addressing the data collection problem for deep learning-based cellular localization systems. The basic idea is to generate synthetic data that reflects the typical pattern of the wireless data as observed from a small collected dataset. Evaluation of the proposed data augmentation techniques using different Android phones in a cellular localization case study shows that we can enhance the performance of the localization systems in both indoor and outdoor scenarios by 157% and 50.5%, respectively. This highlights the promise of the proposed techniques for enabling deep learning-based localization systems.

Abstract In this article we propose a new single‐source shortest‐path algorithm that achieves the same O(n · m) time bound as the Bellman‐Ford‐Moore algorithm but outperforms it and other state‐of‐the‐art algorithms in many cases in practice. Our claims are supported by experimental evidence.

Recent years have witnessed fast growth in outdoor location-based services. While GPS is considered a ubiquitous localization system, it is not supported by low-end phones, requires direct line of sight to the satellites, and can drain the phone battery quickly.

WiFi fingerprinting is one of the mainstream technologies for indoor localization. However, it requires an initial calibration phase during which the fingerprint database is built manually by site surveyors. This process is labour intensive, tedious, and needs to be repeated with any change in the environment. While a number of recent systems have been introduced to reduce the calibration effort through RF propagation models and/or crowdsourcing, these still have some limitations. Other approaches use the recently developed iBeacon technology as an alternative to WiFi for indoor localization. However, these beacon-based solutions are limited to a small subset of high-end phones.