This article presents a novel methodology for predicting wireless signal propagation using ray-tracing algorithms, and visualizing signal variations in situ by leveraging Augmented Reality (AR) tools. The proposed system performs a special type of spatial mapping, capable of converting a scanned indoor environment to a vector facet model. A ray-tracing algorithm uses the facet model for wireless signal predictions. Finally, an AR application overlays the signal strength predictions on the physical space in the form of holograms. Although some indoor reconstruction models have already been developed, this paper proposes an image to a facet algorithm for indoor reconstruction and compares its performance with existing AR algorithms, such as spatial understanding that are modified to create the required facet models. In addition, the paper orchestrates AR and ray-tracing techniques to provide an in situ network visualization interface. It is shown that the accuracy of the derived facet models is acceptable, and the overall signal predictions are not significantly affected by any potential inaccuracies of the indoor reconstruction. With the expected increase of densely deployed indoor 5G networks, it is believed that these types of AR applications for network visualization will play a key role in the successful planning of 5G networks.
Wild-type transthyretin amyloid cardiomyopathy is an under-recognized cause of heart failure. Pfizer previously developed a machine learning model that performed well in identifying wild-type transthyretin amyloid cardiomyopathy vs. nonamyloid heart failure. However, challenges exist when introducing machine learning applications into healthcare, mainly due to restrictions on sharing patient data. This requires the triggering and execution of the model outside the developer’s infrastructure using hosts with diverse information technology capabilities. With these barriers in mind, we investigated architectural designs to facilitate the delivery of the model to the customer. We considered manageability and scalability of the model, the host’s information technology maturity, maintenance of patient data privacy, and protection of Pfizer’s intellectual property. A container-based design, wherein the application is shipped as a container image to a third-party platform, fulfilled these criteria and was piloted on a platform hosted by Philips. Objectives of this pilot included defining/testing the architectural design and technical parameters for sharing the container image, creating a scalable and modular framework to manage multiple applications on different third-party platforms, and exploring a communication pattern based on clinical decision support Hooks/Cards and representational state transfer calls within the Philips platform. Implementing the model may enable earlier identification and treatment of wild-type transthyretin amyloid cardiomyopathy, and learnings from this pilot may lead to improved delivery of other machine learning models to healthcare providers, thereby increasing utilization. This article also presents an overview of architectural designs that may help others to adopt new methodologies and ideas for machine learning model commercialization.
This work presents a novel approach that adopts content caching techniques towards reducing computation and communication costs of Augmented Reality (AR) services. The application scenario under investigation assumes an environment of static objects, each one associated to a holographic content. The goal is to devise practical low-overhead methods so as to reduce the amount of resources above that are needed for the most resource-demanding AR process, namely object recognition. The proposed method is based on caching images using a combination of metrics to rank them such as: (i) an object popularity index which favours objects that are most probable to be requested for recognition, (ii) the percentage of times when the object label has been encountered in the past, (iii) the probability that an image is similar enough with already encountered past images with the same label. The aforementioned image caching method drastically reduces database searches and returns the matched object that satisfies the needs of object recognition. We also devise a binary decision operator that initiates the object recognition process only upon comparison of spatial data of the AR device with the targeted object. The resulting performance is measured using a client-server architecture and components such as Wireshark, Unity Profiler, and Python. For our proposed architecture we deploy an edge server to satisfy the demands of the AR service. Results indicate that the proposed methods can significantly reduce both the computational resources and the induced network traffic, thus improving user experience.
A feedback-based architecture is presented for the distribution grid which enables the use of Machine Learning (ML) techniques for various applications, including Dynamic Voltage Optimization (DVO) and Demand Response (DR). In this architecture, sensor devices are resident on the distribution grid and therefore have a unique awareness of multiple system parameters. This enables the use of ongoing ML techniques for implementation of critical applications in the Smart Grid. Monitoring devices are placed at the endpoints and monitoring/control devices are placed along the power line on various types of grid-resident systems. Because the devices are grid-resident and interact directly with other devices on the same physical link, applications such as ML-assisted DVO can be targeted with very high confidence.