Deployment Of Monitoring Research Articles

Optimizing Distributed Network Monitoring for NFV Service Chains

Efficient and coordinated network monitoring is key to realize the full potential of software-based network infrastructures. In this context, coordinated network monitoring refers to the ability of i) collecting network traffic in a flexible way (i.e., along the path and with different granularities), ii) placing monitoring sinks effectively, and iii) forward collected network traffic efficiently to monitoring sinks. In this letter, we argue that the coordinated network monitoring is another cornerstone for delivering the best-of-breed out of network function virtualization (NFV) infrastructures and an essential step toward fully customized monitoring deployments. We introduce the distributed network monitoring (DNM) problem and propose an optimization model. By coordinating the placement of monitoring sinks and the forwarding of network monitoring traffic, DNM allows finer-grained network monitoring with limited overheads. Results indicate that the DNM outperforms the state-of-the-art telemetry solution by a factor of 3, and reduces of up to 80% in network monitoring traffic that is routed through the network infrastructure.

HS2 railway embankment monitoring: effect of soil condition on underground signals

High speed rail demands precision structural health monitoring in shallow underground environments near embankments. Traditional low frequency wireless channels can communicate with underground sensors, but its large antenna elements are prone to damage from geological stress. In this case study paper, we design a higher frequency compact system and experimentally characterize its performance in-situ, in different soil and moisture conditions that are representative of UK soil conditions near the planned HS2 rail site and weather conditions. Accurate 3D electromagnetic simulation results are also shown to support experimental results and a pathloss model sensitive to soil conditions is developed to inform upcoming high speed rail embankment monitoring deployment. The multi-disciplinary findings presented will also directly inform the depth and data reliability of current high speed rail sensor deployment.

Polymer/Metal−Organic Framework Composite Sensors for Gas Detection

Methane (CH4), the primary composition of the natural gas, is a colorless and odorless gas with a melting point of -184 ºC and a boiling point of -164 ºC. It is considered as a greenhouse gas with a global warming potential far greater than CO2 per molecule, which could significantly contribute to undesired global warming through even moderate amounts of leakage from natural gas recovery and transportation, and can become dangerously explosive for concentrations above the lower explosive limit. Given the fact that the U.S. has nearly 3 million miles of natural gas pipelines, for both safety and environmental concerns, there is an interest in monitoring pipelines for integrity and leak detection. Unfortunately, current sensing technologies do not meet all the practical requirements for widespread deployment and accurate measurement of low level methane leaks or in-pipe gas composition monitoring. Thus, the development of advanced gas sensing devices which are of sufficiently low cost for ubiquitous deployment both internal and external to pipelines, and which can be interrogated remotely over long distances would be highly beneficial. In order to meet the practical requirements of advanced sensor technology for natural gas infrastructure applications, our proposed research seeks to develop a series of polymer/ metal−organic frameworks (MOF) composite thin coating films. The concept targets to exploit the excellent gas adsorption capability and CH4/N2 and CO2/N2 selectivity of highly engineered MOF materials, while a polymer binder is used to keep the well-dispersed mixture composites on the optical fiber surface. The list of possible candidates includes a wide range of polymer/MOF combinations and we herein report the well-dispersed Fe(Pyz)Ni(CN)4 MOF particles blended with the dilute Poly(styrene-butadiene-styrene) (SBS) solution. The suspension is dip-coated onto the optical fiber surface to form a thin layer of film (in nm scale). Unlike pure MOF-based thin coating films which typically rely on time-consuming and complex deposition process, e.g. layer-by-layer technique, this rapid and facile deposition approach offers us the opportunity to efficiently cast materials on long-length (in kilometer scale) distributed platform with a wide range of different candidate MOFs and with high film uniformity and quality. The gas sensing response of the coated fiber is evaluated at room temperature with the transmittance (%T) of the sensor monitored in various gases. When gas molecules are adsorbed by the sensing layer, the refractive index is modified and is recorded by the spectrometer to be correlated with the analyte of interest (Figure 1). A monotonic increase of gas sensing performance is observed by the increase of MOF content in the composites (Figure 1 inset), and is attributed to the enhanced gas adsorption capacity of the MOF matrix. Moreover, a variety of characterizations are performed to investigate the physical/chemical/morphological properties of the polymer/MOF composites and unravel the fundamental understanding of the sensing mechanism. Figure 1. Gas sensing data of a pure SBS polymer and SBS/MOF coated optical fiber. Sensing CH4 and CO2 by monitoring the change in light transmission at 459 nm for SBS/MOF composite and 1944 nm for pure SBS. The inset figure shows the pure CH4 and CO2 sensing response (T%) as a function of MOF wt%. Figure 1

Can We Use Departure from Natural Co-Variance Relationships for Monitoring of Offshore Carbon Storage Integrity?

Carbon capture with offshore storage may take place at various geographical locations, characterized by diverse physical and biogeochemical properties and dynamics of the overlying water. In order to ensure storage integrity, baseline conditions must be carefully assessed for each potential storage area, which will allow design and deployment of optimal monitoring and sampling programs and establish appropriate site-specific criteria for anomaly detection, to allow timely reaction and necessary remedial measures. Within this paper, we assess applicability of using outputs of coupled hydrodynamic-biogeochemical models for the selection of appropriate variables to describe baseline variability and, consequently, strategies for the following monitoring. Via application of multivariate linear regression we identify combinations of modelled variables that best predict variability in pCO2 at a location corresponding to the potential storage site at Goldeneye Field in the Central North Sea. Although some variable pairs better predict pCO2 variability, we focus on a combination of oxygen saturation and silicate, as variables that can potentially be frequently and accurately monitored over long periods. In this work we employ highly simplified leakage scenarios to highlight the accuracy of baseline characterization and implications for establishment of thresholds for anomaly detection in highly dynamic marine environments. We conclude that hydrodynamic-biogeochemical models are invaluable tools for informing cost-effective monitoring strategies regarding the optimal number and combination of parameters surveyed and for establishing appropriate anomaly criteria for each potential storage location.

Successive landsliding and damming of the Jinsha River in eastern Tibet, China: prime investigation, early warning, and emergency response

Two successive landslides within a month started in October 11, 2018, and dammed twice the Jinsha River at the border between Sichuan Province and Tibet in China. Both events had potential to cause catastrophic flooding that would have disrupted lives of millions and induced significant economic losses. Fortunately, prompt action by local authorities supported by the deployment of a real-time landslide early warning system allowed for quick and safe construction of a spillway to drain the dammed lake. It averted the worst scenario without loss of life and property at least one order of magnitude less to what would have been observed without quick intervention. Particularly, the early warning system was able to predict the second large-scale slope failure 24 h in advance, along with minor rock falls during the spillway construction, avoiding false alerts. This paper presents the main characteristics of both slope collapses and damming processes, and introduces the successful landslide early warning system. Furthermore, we found that the slope endured cumulative creeping displacements of > 40 m in the past decade before the first event. Twenty-five meter displacement occurred in the year immediately before. The deformation was measured by the visual interpretation of multitemporal satellite images, which agrees with the interferometry synthetic aperture radar (InSAR) measurement. If these had been done before the emergency, economic losses could have been reduced further. Therefore, our findings strengthen the case for the deployment of systematic monitoring of potential landslide sites by integrating earth observation methods (i.e., multitemporal satellite or UAV images) and in situ monitoring system as a way to reduce risk. It is expected that this success story can be replicated worldwide, contributing to make our society more resilient to landslide events.

Drosophila suzukii (Diptera: Drosophilidae) distribution modelling improves our understanding of pest range limits

Spatial distributions models (SDM) are often used in invasive pest management to understand current and potential distribution. Using data on the well-studied spotted wing drosophila as a model, we compared distribution patterns of the range-limit with commonly applied correlative and mechanistic models. Correlative models risk underestimation whereas simple mechanistic models provide overestimated range predictions, although using both approaches for the spotted wing drosophila improved range-limit predictions. Model choice when dealing with pests is central to the accurate identification of invasive species limit range and consequently for the deployment of monitoring and early detection programs.

A scalable cloud-based cyberinfrastructure platform for bridge monitoring

Cloud computing is a computing paradigm wherein computing resources, such as servers, storage and applications, can be provisioned and accessed in real time via advanced communication networks. In the era of Internet of Things (IoT) and big data, cloud computing has been widely developed in many industrial applications involving large volume of data. Appropriate use of cloud computing infrastructure can enhance the long-term deployment of a structural health monitoring (SHM) system which would incur significant amount of data of different types. This paper presents a cloud-based cyberinfrastructure platform designed to support bridge monitoring. The cyberinfrastructure platform enables scalable management of SHM data and facilitates effective information sharing and data utilisation. A cloud-based platform comprises of virtual machines, distributed database and web servers. The peer-to-peer distributed database architecture provides a scalable and fault-tolerant data management system. Platform-neutral web services designed in compliant with the Representational State Transfer (REST) standard enables easy access to the cloud resources and SHM data. For data interoperability, a bridge information model for bridge monitoring applications is adopted. For demonstration, the scalable cloud-based platform is implemented for the monitoring of bridges along the I-275 corridor in the State of Michigan. The results show that the cloud-based cyberinfrastructure platform can effectively manage the sensor data and bridge information and facilitate efficient access of the data as well as the bridge monitoring software services.

A Survey on Sensor Calibration in Air Pollution Monitoring Deployments

Air pollution is a major concern for public health and urban environments. Conventional air pollution monitoring systems install a few highly accurate, expensive stations at representative locations. Their sparse coverage and low spatial resolution are insufficient to quantify urban air pollution and its impacts on human health and environment. Advances in low-cost portable air pollution sensors have enabled air pollution monitoring deployments at scale to measure air pollution at high spatiotemporal resolution. However, it is challenging to ensure the accuracy of these low-cost sensor deployments because the sensors are more error-prone than high-end sensing infrastructures and they are often deployed in harsh environments. Sensor calibration has proven to be effective to improve the data quality of low-cost sensors and maintain the reliability of long-term, distributed sensor deployments. In this paper, we review the state-of-the-art low-cost air pollution sensors, identify their major error sources, and comprehensively survey calibration models as well as network recalibration strategies suited for different sensor deployments. We also discuss limitations of exiting methods and conclude with open issues for future sensor calibration research.

S85. Automated quantitative EEG reactivity testing using actigraphy monitoring in cardiac arrest coma prognostication

Introduction EEG background reactivity is a strong predictor of coma recovery after cardiac arrest. The current clinical value of EEG background reactivity testing is limited by inadequate inter-rater expert agreement, few number of daily assessments, and unsuitability for quantitative tracking by visual review. We hypothesized that a quantitative EEG reactivity method using actigraphy-triggered events from a wrist-worn wearable can predict long-term functional outcome in comatose cardiac arrest patients treated with targeted temperature management. Methods We prospectively recorded clinical, actigraphy, and EEG data of comatose cardiac arrest patients managed with targeted temperature management at a single U.S. academic hospital. Continuous actigraphy data obtained from a wrist-worn wearable (Affectiva 3) was synchronized to continuous EEG data for up to 96 h after cardiac arrest. Change in actigraphy across time was used as a surrogate of bedside external physical stimulation (i.e. actigraphy-triggered event). Our quantitative EEG method evaluated changes in EEG spectra, entropy, and frequency features during 30 s before and 30 s after each actigraphy-triggered event (46 EEG features used). Actigraphy threshold and EEG window duration (pre and post actigraphy-triggered events) were utilized as hyperparameters in the model. In addition to actigraphy-triggered quantitative EEG background reactivity, EEG reactivity was scored visually as present or absent once daily by a single expert electroencephalographer during clinical care. Good outcome was defined as Cerebral Performance Category of 1–2 at six months. Classification was carried out using extreme learning machine and leave-one-subject-out cross validation. Final outcome class prediction for each individual subject was determined using simple majority voting across all actigraphy-triggered events. Results Ten subjects were monitored. Mean age was 60 (standard deviation (SD) ± 14.6) years, 10% were female, 50% had a shockable rhythm, and 40% had good outcome at six months. A total of 19 (SD ± 8.9) actigraphy-triggered events per patient were detected. Best outcome prediction performance was achieved using 50-s-long EEG window (25-s pre and 25-s post each actigraphy-triggered event). The quantitative EEG reactivity method correctly predicted good outcome in 80% of cases (sensitivity and specificity 80%). Expert rating using once daily visual EEG reactivity assessments had 60% accuracy for good outcome prediction. Conclusion Quantitative EEG reactivity assessments using actigraphy-based triggered events is feasible and may support long-term outcome prediction in cardiac arrest coma. Integration of EEG and wearable sensors using machine learning methods might facilitate the deployment of dynamic multi-modal patient monitoring in the intensive care unit environment.

Parameter-Invariant Monitor Design for Cyber–Physical Systems

The tight interaction between information technology and the physical world inherent in cyber–physical systems (CPS) can challenge traditional approaches for monitoring safety and security. Data collected for robust CPS monitoring is often sparse and may lack rich training data describing critical events/attacks. Moreover, CPS often operate in diverse environments that can have significant inter/intra-system variability. Furthermore, CPS monitors that are not robust to data sparsity and inter/intra-system variability may result in inconsistent performance and may not be trusted for monitoring safety and security. Towards overcoming these challenges, this paper presents recent work on the design of parameter-invariant (PAIN) monitors for CPS. PAIN monitors are designed such that unknown events and system variability minimally affect the monitor performance. This work describes how PAIN designs can achieve a constant false alarm rate (CFAR) in the presence of data sparsity and intra/inter system variance in real-world CPS. To demonstrate the design of PAIN monitors for safety monitoring in CPS with different types of dynamics, we consider systems with networked dynamics, linear-time invariant dynamics, and hybrid dynamics that are discussed through case studies for building actuator fault detection, meal detection in type I diabetes, and detecting hypoxia caused by pulmonary shunts in infants. In all applications, the PAIN monitor is shown to have (significantly) less variance in monitoring performance and (often) outperforms other competing approaches in the literature. Finally, an initial application of PAIN monitoring for CPS security is presented along with challenges and research directions for future security monitoring deployments.

Rapid deployment of a WSN on the Clifton Suspension Bridge, UK

The Clifton Suspension Bridge is an emblematic structure in Bristol, UK. In this paper, a rapid deployment of a structural health monitoring (SHM) system for short-term monitoring is described. The system, deployed in a single day, integrates wireless SHM and open-source data management systems to gather valuable information about the bridge use (loading). The deployed system can be used to inform structural response models as well as studies for traffic engineering purposes. The use of open-source software was critical to the successful deployment.

Mechanical Performance Monitoring for Prestressed Concrete Piles Used in a Newly-Built High-Piled Wharf in a Harbor with Fiber Bragg Grating Sensor Technology When Pile Driving

Mechanical performance monitoring of civil infrastructure using fiber Bragg grating (FBG) sensors has received significant public attention in recent years. However, there is currently little research on the mechanical performance monitoring of piles used in high-piled wharfs in coastal ports during pile driving using the FBG sensor technique. Based on the properties of precast prestressed concrete piles used in high-piled wharfs in coastal ports and servicing seawater environments, and the benefits of FBG sensors, the mechanical performance monitoring for precast prestressed concrete piles used in a newly-built high-piled wharf in the Tianjin Port of China is devised and deployed with the FBG sensor technique. To conduct performance monitoring of the precast prestressed concrete pile, a state-of-the-art FBG strain sensor, which is less thermosensitive and does not require temperature compensation, was used to monitor the strain status of different locations of the pile. In one pile, three of this kind of strain sensor were set near the head, middle and tip of the pile, and one FBG angle sensor was set near the head of the pile to measure the dip angle of the pile. During the testing, data were recorded for all of the details of the pile driving process. According to the data analysis, it is clear that the compressive strain at the middle of the pile during driving is larger than that near the head and tip of the pile. Therefore, the middle of the prestressed concrete pile is the key location that should be preferentially monitored during pile driving. Meanwhile, when the hammer impacts the pile continuously, the obvious tension strain at the tip of the pile increases and the maximum dynamic tension strain reaches 56 με, which approaches the tension ultimate strain. This occurs because the frictional resistance of soil is small in the middle of the pile when the tip meets the significant supporting soil layer. This study can provide a reference for the mechanical performance monitoring deployment of precast prestressed concrete piles used in high-piled wharf structures in coastal ports.

Wireless Sensor Network Based Smart Grid Communications: Challenges, Protocol Optimizations, and Validation Platforms

Smart grids, the next generation of electric grids, require the deployment of sophisticated monitoring and control systems to enhance their operational efficiency. Wireless sensor networks (WSNs) have been considered as a promising communication technology for the monitoring and control of smart grid operation. They bring significant advantages such as, rapid deployment, low cost and scalability. However, the deployment of WSNs in smart grids brought new challenges mainly due to the electric grid features. Consequently, traditional WSN communication protocols have been shown inadequate and several recent research efforts were dedicated for their optimization. This paper provides a comprehensive survey on related literature, discusses the still-open research issues, and identifies the most common validation platforms for experimenting WSN communications in smart grid. We believe this survey will pave the way for the research community to (i) understand important concepts related to WSN-based smart grid communications, (ii) identify gaps and make valuable contributions in this timely and exiting field and (iii) choose the convenient experimental platform for the validation of proposed solutions.

Wireless Magnetic Sensor Network for Road Traffic Monitoring and Vehicle Classification

Abstract Efficiency of transportation of people and goods is playing a vital role in economic growth. A key component for enabling effective planning of transportation networks is the deployment and operation of autonomous monitoring and traffic analysis tools. For that reason, such systems have been developed to register and classify road traffic usage. In this paper, we propose a novel system for road traffic monitoring and classification based on highly energy efficient wireless magnetic sensor networks. We develop novel algorithms for vehicle speed and length estimation and vehicle classification that use multiple magnetic sensors. We also demonstrate that, using such a low-cost system with simplified installation and maintenance compared to current solutions, it is possible to achieve highly accurate estimation and a high rate of positive vehicle classification.

Cost and benefit estimates of partially-automated vehicle collision avoidance technologies

Many light-duty vehicle crashes occur due to human error and distracted driving. Partially-automated crash avoidance features offer the potential to reduce the frequency and severity of vehicle crashes that occur due to distracted driving and/or human error by assisting in maintaining control of the vehicle or issuing alerts if a potentially dangerous situation is detected. This paper evaluates the benefits and costs of fleet-wide deployment of blind spot monitoring, lane departure warning, and forward collision warning crash avoidance systems within the US light-duty vehicle fleet. The three crash avoidance technologies could collectively prevent or reduce the severity of as many as 1.3 million U.S. crashes a year including 133,000 injury crashes and 10,100 fatal crashes. For this paper we made two estimates of potential benefits in the United States: (1) the upper bound fleet-wide technology diffusion benefits by assuming all relevant crashes are avoided and (2) the lower bound fleet-wide benefits of the three technologies based on observed insurance data. The latter represents a lower bound as technology is improved over time and cost reduced with scale economies and technology improvement. All three technologies could collectively provide a lower bound annual benefit of about $18 billion if equipped on all light-duty vehicles. With 2015 pricing of safety options, the total annual costs to equip all light-duty vehicles with the three technologies would be about $13 billion, resulting in an annual net benefit of about $4 billion or a $20 per vehicle net benefit. By assuming all relevant crashes are avoided, the total upper bound annual net benefit from all three technologies combined is about $202 billion or an $861 per vehicle net benefit, at current technology costs. The technologies we are exploring in this paper represent an early form of vehicle automation and a positive net benefit suggests the fleet-wide adoption of these technologies would be beneficial from an economic and social perspective.

A Methodology for Fast Deployment of Condition Monitoring and Generic Services Platform Technological Design

Maintenance is a research field that has recently been gaining importance in business and where the study and development of monitoring and predictive technologies has been very active, as the role of these technologies is key in enabling predict and prevent maintenance strategies. Moreover, by means of monitoring features of processes and components, an impact in lifecycle value can be achieved. However, challenges remain in structuring the condition monitoring offer and the technological platform due, in particular, to the variety of potential domains of application, the characteristics of the existing information and the final goals of the monitoring activities. These challenges may impact in the deployment time of a condition monitoring solution. In order to limit these challenges, a methodology for fast deployment of condition monitoring and a technological service platform is presented. The methodology has been obtained from research and analysis of several use cases in the context of product-service systems. The focus is on methodological and technological results, which are presented in a general manner such that they can be applicable to the deployment of condition monitoring and services in various domains. Finally, application of the methodology is presented in two different scenarios.

Roadmap for the deployment of WAMPAC in the future GB power system

This study presents a specific roadmap for the deployment of a wide area monitoring, protection, and control (WAMPAC) system in the future Great Britain (GB) power system. The specific roadmap is designed based on a generic roadmap, which draws on international experience with the development of WAMPAC systems, and a study of the changes in the GB power system. Based on this, the proposed specific roadmap identifies wide area oscillation monitoring as a key short term goal for the GB WAMPAC system and wide area oscillation control as a key long term goal for the GB WAMPAC system. To verify that this short term goal can be delivered simulations of a complete model of the GB power system that demonstrate that a new method based on a Newton Type algorithm will be more than capable of providing accurate oscillation monitoring in the GB WAMPAC system. Furthermore, for the same complete GB model it is demonstrated that the West coast CSC HVDC link can be used to create a powerful tool for oscillation damping that, with further research, should be capable of delivering the long-term goal of wide area oscillation control as part of the specific GB WAMPAC roadmap.

Automated in-core image generation from video to aid visual inspection of nuclear power plant cores

Inspection and monitoring of key components of nuclear power plant reactors is an essential activity for understanding the current health of the power plant and ensuring that they continue to remain safe to operate. As the power plants age, and the components degrade from their initial start-of-life conditions, the requirement for more and more detailed inspection and monitoring information increases. Deployment of new monitoring and inspection equipment on existing operational plant is complex and expensive, as the effect of introducing new sensing and imaging equipment to the existing operational functions needs to be fully understood. Where existing sources of data can be leveraged, the need for new equipment development and installation can be offset by the development of advanced data processing techniques.This paper introduces a novel technique for creating full 360° panoramic images of the inside surface of fuel channels from in-core inspection footage. Through the development of this technique, a number of technical challenges associated with the constraints of using existing equipment have been addressed. These include: the inability to calibrate the camera specifically for image stitching; dealing with additional data not relevant to the panorama construction; dealing with noisy images; and generalising the approach to work with two different capture devices deployed at seven different Advanced Gas Cooled Reactor nuclear power plants. The resulting data processing system is currently under formal assessment with a view to replacing the existing manual assembly of in-core defect montages. Deployment of the system will result in significant time savings on the critical outage path for the plant operator and will allow improved visualization of the surface of the inside of fuel channels, far beyond that which can be gained from manually analysing the raw video footage as is done at present.

Structural Health Monitoring in Incrementally Launched Steel Bridges: Patch Loading Phenomena Modeling

Abstract In this paper, a realistic nonlinear 3D simulation of an incrementally launched steel bridge girder is presented. The numerical simulation accounts for three sources of nonlinearity: geometry, material and boundary conditions. For the sake of depicting the capabilities of the presented numerical model in structural verifications, the study is focused on the patch loading field, a structurally complex phenomenon. Patch loading (or concentrated loading) is one of the most typically encountered structural verifications on incrementally launched steel I-girders. The presented realistic simulation is based upon an experimentally calibrated numerical model and may provide relevant information at both design and construction stages. For the former, the predictive capabilities of the model for inferring the potential failure due to patch loading are depicted. For the latter, the results obtained are displayed in a way that may be useful for planning a Structural Health Monitoring (SHM) deployment aimed at controlling the patch loading-related phenomena in incrementally launched steel plate girders.

(244) Stanford Patient Experience Questionnaire (SPEQ): Machine-mediated classification of patient experience feedback using natural language processing

Patient experience, an emerging marker of healthcare quality to policy makers and payers, is an area where Pain Medicine has historically underperformed. Real-time monitoring of patient feedback may be a critical tool to optimize the patient experience while eliminating the need for manual intervention. Here we leverage the Collaborative Health Outcomes Information Registry (CHOIR) platform and the Stanford Patient Experience Questionnaire (SPEQ) and develop a machine-mediated automatic classification of patient free-text feedback in terms of the sentiment expressed. SPEQ measures patient experience with 20 items in 7 touch points plus one free-text item. Over a three-week period at an academic multidisciplinary pain center, 628 patients were given SPEQ post-visit, of whom 123 (20%) completed the questionnaire, of whom 63 (51%) provided free-text feedback. The length of the feedback ranged from 2 to 513 words, with an average of 81.5 and SD of 105.8 words. We performed manual coding in terms of 1) aspects of patient experience referenced, 2) sentiment expressed for each aspect, and 3) overall sentiment. Next we performed natural language processing (NLP) using recursive neural network (Socher et al 2013), which rates each sentence as having Positive, Neutral, or Negative sentiment. In order to derive an overall rating of passage-level sentiment, we performed a receiver-operating curve analysis. The optimal cut-off for classifying the feedback as Negative is a threshold of 68.5% or more of the sentences being Negative. There is moderate agreement between machine-mediated rating with single-rater coding, with Kappa of 0.57 and AUC of 0.79. In summary, we have developed an algorithm that enables automatic classification of patient feedback. Large-volume and real-time processing of patient feedback enables organization-wide deployment of patient experience monitoring with low cost and resource burden. Future work will focus on optimizing this algorithm using training data from additional patient feedback. Patient experience, an emerging marker of healthcare quality to policy makers and payers, is an area where Pain Medicine has historically underperformed. Real-time monitoring of patient feedback may be a critical tool to optimize the patient experience while eliminating the need for manual intervention. Here we leverage the Collaborative Health Outcomes Information Registry (CHOIR) platform and the Stanford Patient Experience Questionnaire (SPEQ) and develop a machine-mediated automatic classification of patient free-text feedback in terms of the sentiment expressed. SPEQ measures patient experience with 20 items in 7 touch points plus one free-text item. Over a three-week period at an academic multidisciplinary pain center, 628 patients were given SPEQ post-visit, of whom 123 (20%) completed the questionnaire, of whom 63 (51%) provided free-text feedback. The length of the feedback ranged from 2 to 513 words, with an average of 81.5 and SD of 105.8 words. We performed manual coding in terms of 1) aspects of patient experience referenced, 2) sentiment expressed for each aspect, and 3) overall sentiment. Next we performed natural language processing (NLP) using recursive neural network (Socher et al 2013), which rates each sentence as having Positive, Neutral, or Negative sentiment. In order to derive an overall rating of passage-level sentiment, we performed a receiver-operating curve analysis. The optimal cut-off for classifying the feedback as Negative is a threshold of 68.5% or more of the sentences being Negative. There is moderate agreement between machine-mediated rating with single-rater coding, with Kappa of 0.57 and AUC of 0.79. In summary, we have developed an algorithm that enables automatic classification of patient feedback. Large-volume and real-time processing of patient feedback enables organization-wide deployment of patient experience monitoring with low cost and resource burden. Future work will focus on optimizing this algorithm using training data from additional patient feedback.