Engineering Papers⌕ Search

SEARCH · Engineering Papers

Results for “Inference”

Search indexed NASA NTRS and DOE OSTI research on propulsion, heat transfer, battery materials and energy systems. Follow report and document links to the original sources.

Quote a phrase for an exact phrase match. Source license links do not imply unrestricted reuse.

At least 469 records · Page 26

Genetic Network Inference: From Co-Expression Clustering to Reverse Engineering

Advances in molecular biological, analytical, and computational technologies are enabling us to systematically investigate the complex molecular processes underlying biological systems. In particular, using high-throughput gene expression assays, we are able to measure the output of the gene regulatory network. We aim here to review datamining and modeling approaches for conceptualizing and unraveling the functional relationships implicit in these datasets. Clustering of co-expression profiles allows us to infer shared regulatory inputs and functional pathways. We discuss various aspects of clustering, ranging from distance measures to clustering algorithms and multiple-duster memberships. More advanced analysis aims to infer causal connections between genes directly, i.e., who is regulating whom and how. We discuss several approaches to the problem of reverse engineering of genetic networks, from discrete Boolean networks, to continuous linear and non-linear models. We conclude that the combination of predictive modeling with systematic experimental verification will be required to gain a deeper insight into living organisms, therapeutic targeting, and bioengineering.

Dhaeseleer, Patrik↗

Inferring Cloud Properties from Passive Solar Sensors

The most common approach for retrieving cloud optical depth from ground-based observations uses downwelling fluxes measured by pyranometers and Multi-Filter Rotating Shadowband Radiometers (MFRSR). The key element in both retrieval techniques is the one-to-one mapping of the "observed" fluxes into cloud optical depth through plane-parallel radiative transfer. Both methods are expected to work well only for completely overcast clouds giving an effective optical depth for the whole sky. To infer cloud optical depth locally one can assume to use a narrow-field-of view radiometer that measures radiances instead of fluxes. However, lack of one-to-one relationships between radiance and cloud optical depth, prevents the direct use of radiances either. Here we discuss a new method for retrieving cloud optical depth from ground-based measurements of zenith radiance in the RED and near infrared (NIR) spectral regions. In addition to an algebraic combinations of spectral radiances such as Normalized Difference Cloud Index (NDCI) that largely removes nonuniqueness and the radiative effects of cloud inhomogeneity but can result in poor retrievals due to its insensitivity to cloud fraction, both RED and NIR radiances as points on the "RED vs. NIR" plane are proposed to be used for retrieval. The proposed retrieval method is applied to a multi- channel sunphotometer (as a part of AERONET) measurements at the ARM site in Oklahoma. The results of retrieval are compared with the ones from Microwave Radiometer (MWR) and MFRSR. In addition to cloud optical depth, the new retrieval method also infers a "radiatively effective" cloud fraction.

Marshak, A.↗

F-OWL: An Inference Engine for Semantic Web

Understanding and using the data and knowledge encoded in semantic web documents requires an inference engine. F-OWL is an inference engine for the semantic web language OWL language based on F-logic, an approach to defining frame-based systems in logic. F-OWL is implemented using XSB and Flora-2 and takes full advantage of their features. We describe how F-OWL computes ontology entailment and compare it with other description logic based approaches. We also describe TAGA, a trading agent environment that we have used as a test bed for F-OWL and to explore how multiagent systems can use semantic web concepts and technology.

Zou, Youyong↗

Inference of Ice Cloud Properties from High-spectral Resolution Infrared Observations

The theoretical basis is explored for inferring the microphysical properties of ice crystal from high-spectral resolution infrared observations. A radiative transfer model is employed to simulate spectral radiances to address relevant issues. The extinction and absorption efficiencies of individual ice crystals, assumed as hexagonal columns for large particles and droxtals for small particles, are computed from a combination of the finite- difference time-domain (FDTD) technique and a composite method. The corresponding phase functions are computed from a combination of FDTD and an improved geometric optics method (IGOM). Bulk scattering properties are derived by averaging the single- scattering properties of individual particles for 30 particle size distributions developed from in situ measurements and for additional four analytical Gamma size distributions for small particles. The non-sphericity of ice crystals is shown to have a significant impact on the radiative signatures in the infrared (IR) spectrum; the spherical particle approximation for inferring ice cloud properties may result in an overest&ation of the optical thickness and an inaccurate retrieval of effective particle size. Furthermore, we show that the error associated with the use of the Henyey-Greenstein phase function can be as larger as 1 K in terms of brightness temperature for larger particle effective size at some strong scattering wavenumbers. For small particles, the difference between the two phase functions is much less, with brightness temperatures generally differing by less than 0.4 K. The simulations undertaken in this study show that the slope of the IR brightness temperature spectrum between 790-960/cm is sensitive to the effective particle size. Furthermore, a strong sensitivity of IR brightness temperature to cloud optical thickness is noted within the l050-1250/cm region. Based on this spectral feature, a technique is presented for the simultaneous retrieval of the visible optical thickness and effective particle size from high spectral resolution infrared data under ice cloudy con&tion. The error analysis shows that the uncertainty of the retrieved optical thickness and effective particle size has a small range of variation. The error for retrieving particle size in conjunction with an uncertainty of 5 K in cloud'temperature, or a surface temperature uncertainty of 2.5 K, is less than 15%. The corresponding e m r in the uncertainty of optical thickness is within 5-2096, depending on the value of cloud optical thickness. The applicability of the technique is demonstrated using the aircraft-based High- resolution Interferometer Sounder (HIS) data from the Subsonic Aircraft: Contrail and Cloud Effects Special Study (SUCCESS) in 1996 and the First ISCCP Regional Experiment - Arctic Clouds Experiment (FIRE-ACE) in 1998.

Huang, Hung-Lung↗

Use of Minute-by-Minute Cardiovascular Measurements During Tilt Tests to Strengthen Inference on the Effect of Long-Duration Space Flight on Orthostatic Hypotension

Typical methodology for evaluating the effects of spaceflight on orthostatic hypotension (OH) has been survival analysis of tolerance times from 80 head-up tilt tests. However when scheduled test durations are short, there may not be enough failures to allow survival analysis to adequately estimate and compare the effects of flight phase (e.g. pre-flight, number of days post-flight), flight duration, and their interaction, as well as interactions with effects of interventions or countermeasures. The problem is exacerbated in the presence of a repeated measures design, in which subjects participate in tilt tests during various flight phases. Here we show how it is possible to dramatically improve the efficiency of statistical inference in this setting by making use of the additional information contained in minute-by-minute observations of cardiovascular parameters thought to be reflective of progression towards presyncope during tilt testing. Methods: We retrospectively examined operational tilt test (OTT; 10 -min 80 head-up tilt) data from 20 International Space Station (ISS) and 66 Shuttle astronauts 10 d before launch (L-10), on landing day (R+0) and during recovery (R+1, R+3, R+6-10) depending on the level of participation. Data from 5 ISS astronauts tested on R+0 or R+1 who used non-standard countermeasures were excluded. In addition to OTT survival time, 8 cardiovascular parameters (CP: heart rate, systolic, diastolic, and mean arterial blood pressure, pulse pressure, stroke volume, cardiac output, and total peripheral resistance) that might be predictive of progression towards presyncope were measured every minute of each OTT. Statistical analysis was predicated on a two ]stage model of causation. In the first stage, flight duration and time from landing affect the astronauts' degree of OH, which is manifested in the time trends and variation of the above CPs during OTTs. In the second stage, the behavior of these parameters directly affects OTT survival time. Actual analysis proceeded in the opposite direction. First we identified those CPs or linear combinations that best predicted OTT survival regardless of what spaceflight conditions led to OTT completion or presyncope. From these, we calculated a summary statistic (one per OTT) that best predicted survival. We then used mixed ]model regression analysis to relate changes in the summary statistic to flight phase and duration. Inference on the effects of phase, duration, and their interaction on OH follows directly from this second analysis. Results: A linear combination (W) of diastolic blood pressure (DBP) and stroke volume (SV) was found to be the best predictor of OTT survival using the complete data set of minute-by-minute observations of CPs for each OTT. Furthermore, the log-transformed standard deviation of W (Z = log SW) was found to be a strong predictor of survival in the reduced data set consisting of one observation per OTT. In other words, this measure of variability of W during an OTT was the best indicator of whether or not the subject could complete the 10-min test, with higher variability (i.e. higher values of Z) being associated with greater probability of failure. In the mixed-model regression analysis where Z was now treated as a outcome with flight phase and duration groups (ISS and STS) as predictors, we found that there was a significantly more variability in W (higher values of Z) for both groups at R+0, but with no evidence of an interaction until R+3, when the ISS group still had inflated variability, but not the STS group. Conclusions: Variability of the cardiovascular index W recovers more slowly after long-compared to short-duration spaceflight. Since high variability of W has also been shown to be predictive of OTT failure, a primary manifestation of OH, a logical conclusion is that recovery from OH also is slower after long-duration compared to short-duration spaceflights.

Feiveson, Alan H.↗

CALIPSO-Inferred Aerosol Direct Radiative Effects: Bias Estimates Using Ground-Based Raman Lidars

Observational constraints on the change in the radiative energy budget caused by the presence of aerosols, i.e. the aerosol direct radiative effect (DRE), have recently been made using observations from the Cloud- Aerosol Lidar and Infrared Pathfinder Satellite (CALIPSO). CALIPSO observations have the potential to provide improved global estimates of aerosol DRE compared to passive sensor-derived estimates due to CALIPSO's ability to perform vertically-resolved aerosol retrievals over all surface types and over cloud. In this study we estimate the uncertainties in CALIPSO-inferred aerosol DRE using multiple years of observations from the Atmospheric Radiation Measurement (ARM) program's Raman lidars (RL) at mid-latitude and tropical sites. Examined are assumptions about the ratio of extinction-to-backscatter (i.e. the lidar ratio) made by the CALIPSO retrievals, which are needed to retrieve the aerosol extinction profile. The lidar ratio is shown to introduce minimal error in the mean aerosol DRE at the top-of-atmosphere and surface. It is also shown that CALIPSO is unable to detect all radiatively-significant aerosol, resulting in an underestimate in the magnitude of the aerosol DRE by 30â€"50%. Therefore, global estimates of the aerosol DRE inferred from CALIPSO observations are likely too weak.

Thorsen, Tyler↗

CALIPSO-Inferred Aerosol Direct Radiative Effects: Bias Estimates Using Ground-Based Raman Lidars

Observational constraints on the change in the radiative energy budget caused by the presence of aerosols, i.e. the aerosol direct radiative effect (DRE), have recently been made using observations from the Cloud- Aerosol Lidar and Infrared Pathfinder Satellite (CALIPSO). CALIPSO observations have the potential to provide improved global estimates of aerosol DRE compared to passive sensor-derived estimates due to CALIPSO's ability to perform vertically-resolved aerosol retrievals over all surface types and over cloud. In this study we estimate the uncertainties in CALIPSO-inferred aerosol DRE using multiple years of observations from the Atmospheric Radiation Measurement (ARM) program's Raman lidars (RL) at midlatitude and tropical sites. Examined are assumptions about the ratio of extinction-to-backscatter (i.e. the lidar ratio) made by the CALIPSO retrievals, which are needed to retrieve the aerosol extinction profile. The lidar ratio is shown to introduce minimal error in the mean aerosol DRE at the top-of-atmosphere and surface. It is also shown that CALIPSO is unable to detect all radiatively-significant aerosol, resulting in an underestimate in the magnitude of the aerosol DRE by 30-50%. Therefore, global estimates of the aerosol DRE inferred from CALIPSO observations are likely too weak.

Thorsen, Tyler↗

Proprioceptive Inference for Dual-Arm Grasping of Bulky Objects Using RoboSimian

This work demonstrates dual-arm lifting of bulky objects based on inferred object properties (center of mass (COM) location, weight, and shape) using proprioception (i.e. force torque measurements). Data-driven Bayesian models describe these quantities, which enables subsequent behaviors to depend on confidence of the learned models. Experiments were conducted using the NASA Jet Propulsion Laboratory’s (JPL) RoboSimian to lift a variety of cumbersome objects ranging in mass from 7kg to 25kg. The position of a supporting second manipulator was determined using a particle set and heuristics that were derived from inferred object properties. The supporting manipulator decreased the initial manipulator’s load and distributed the wrench load more equitably across each manipulator, for each bulky object. Knowledge of the objects came from pure proprioception (i.e. without reliance on vision or other exteroceptive sensors) throughout the experiments.

Backes, Paul↗

Quantifying Uncertainty in Particle Size Distribution Parameters Inferred from SAGE III/ISS Extinction Spectra

Stratospheric aerosols play key roles in the chemistry and radiation balance of the atmosphere and are a key input parameter for global chemistry and climate models. The degree to which aerosols impact chemistry and radiation balance depends primarily on their microphysical properties such as particle size distribution (PSD). The PSD is a mathematical description of the relative abundance of different sized particles within a sampling volume. If the PSD is accurately known then other key modeling parameters (e.g., surface area density) can be derived. Occultation observations from orbital instruments such as SAGE III/ISS have been used to infer these PSD parameters by inverting the extinction coefficient spectra. However, past efforts failed to address two key issues with this methodology: 1. The measurement uncertainty was not accounted for; 2. They assumed the PSDs to be single-mode only, while “real-world” PSDs are predominantly bi-modal. Accounting for both issues in the retrieval will yield an expanded solution space to the inferred PSD parameters; the question is “by how much?” To address this knowledge gap, we propose to carry out a series of simulations and, for every valid SAGE III/ISS data point, determine the range of PSD parameters that yield extinction spectra that are indistinguishable from the SAGE III/ISS data, within the limits of the reported uncertainty. Further, we will expand the solution space, for the first time, to include bimodal distributions. The results of this work will advance Earth system modeling/prediction capability through identifying the uncertainty of PSD parameter estimates using SAGE III/ISS data. The key benefits of this study over previous studies are twofold: 1. we will provide PSD estimates that include bimodal distributions in the solution space, 2. we will provide an uncertainty estimate for these parameters. The results of this study may be used directly in current and future climate and chemistry models.

SAGE III/ISS↗

Using Machine Learning to Infer Material Properties of Debris Fragments from X-ray Images in the DebriSat Project

The DebriSat project is a collaboration effort with the NASA Orbital Debris Program Office, the U.S. Space Force Space Systems Command Center, The Aerospace Corporation, and the University of Florida. To date, over 200,000 fragments from this ground-based, hypervelocity impact experiment have been collected, and processing is underway to determine their physical characteristics, such as material, shape, color, characteristic length, and average cross-sectional area. The x-ray process is primarily used to identify the location of the fragments and estimated size for extraction, so that these physical characteristics can be assessed. This paper proposes a machine learning-based approach to characterize materials from x-ray images of debris fragments embedded in soft-catch foam used in the DebriSat project. The novel methodology discussed in this paper will highlight the use of x-ray imagery data to characterize these fragments without extraction or a human-in-the-loop. Both supervised and unsupervised machine learning techniques are utilized with this approach to infer the physical parameters of the fragments embedded in the soft-catch foam panels used in the impact experiment based on x-ray images of the foam panels. Additionally, 3D reconstructions of the extracted fragments are created with images taken from two different angles using the structure from motion (SfM) method. The characteristic lengths and shape from the 3D reconstruction, alongside the physical characteristics of the debris, are used in the inference of the material type. To develop and test the approach, a dataset of x-ray images of debris fragments of varying sizes and materials is collected. Supervised learning methods such as convolutional neural networks (CNNs), support vector machines (SVM), decision trees, and random forest classifiers are used due to the high-dimensional feature spaces of the debris and nonlinear decision boundaries for material categorization. Given the limited pre-labeled data of embedded debris materials smaller than 10 mm, unsupervised machine learning techniques such as clustering algorithms and autoencoders are used, in addition to supervised learning methods. The clustering algorithms group similar fragments together based on their physical properties, and autoencoders reduce the dimensionality of the x ray images and extract relevant features. The performance of the proposed approach's is analyzed using a range of statistical methods, including confusion matrices, receiver operating characteristic curves, and precision-recall curves. The results are compared with those obtained using a baseline approach that relies on manual identification and classification of debris fragments. To evaluate the effectiveness of different machine learning methods, statistical tests such as t-tests, ANOVA, and cross-validation are performed, comparing the performance of CNNs, SVMs, clustering algorithms, and autoencoders. Additional analysis needs to be conducted to identify any sources of bias or variability that may affect the results, such as variations in imaging conditions or fragmentation patterns. Other topics explored are limitations, refinements, and the potential use of semi-supervised learning techniques, such as self-training to label unlabeled datasets and co-training using x-ray images taken from two different angles as two different models.

Saik Anam Siam↗

Using Machine Learning to Infer Material Properties of Debris Fragments from X-ray Images in the DebriSat Project

The DebriSat project is a collaboration effort with the NASA Orbital Debris Program Office, the U.S. Space Force Space Systems Command Center, The Aerospace Corporation, and the University of Florida. To date, over 200,000 fragments from this ground-based, hypervelocity impact experiment have been collected, and processing is underway to determine their physical characteristics, such as material, shape, color, characteristic length, and average cross-sectional area. The x-ray process is primarily used to identify the location of the fragments and estimated size for extraction, so that these physical characteristics can be assessed. This paper proposes a machine learning-based approach to characterize materials from x-ray images of debris fragments embedded in soft-catch foam used in the DebriSat project. The novel methodology discussed in this paper will highlight the use of x-ray imagery data to characterize these fragments without extraction or a human-in-the-loop. Both supervised and unsupervised machine learning techniques are utilized with this approach to infer the physical parameters of the fragments embedded in the soft-catch foam panels used in the impact experiment based on x-ray images of the foam panels. Additionally, 3D reconstructions of the extracted fragments are created with images taken from two different angles using the structure from motion (SfM) method. The characteristic lengths and shape from the 3D reconstruction, alongside the physical characteristics of the debris, are used in the inference of the material type. To develop and test the approach, a dataset of x-ray images of debris fragments of varying sizes and materials is collected. Supervised learning methods such as convolutional neural networks (CNNs), support vector machines (SVM), decision trees, and random forest classifiers are used due to the high-dimensional feature spaces of the debris and nonlinear decision boundaries for material categorization. Given the limited pre-labeled data of embedded debris materials smaller than 10 mm, unsupervised machine learning techniques such as clustering algorithms and autoencoders are used, in addition to supervised learning methods. The clustering algorithms group similar fragments together based on their physical properties, and autoencoders reduce the dimensionality of the x ray images and extract relevant features. The performance of the proposed approach's is analyzed using a range of statistical methods, including confusion matrices, receiver operating characteristic curves, and precision-recall curves. The results are compared with those obtained using a baseline approach that relies on manual identification and classification of debris fragments. To evaluate the effectiveness of different machine learning methods, statistical tests such as t-tests, ANOVA, and cross-validation are performed, comparing the performance of CNNs, SVMs, clustering algorithms, and autoencoders. Additional analysis needs to be conducted to identify any sources of bias or variability that may affect the results, such as variations in imaging conditions or fragmentation patterns. Other topics explored are limitations, refinements, and the potential use of semi-supervised learning techniques, such as self-training to label unlabeled datasets and co-training using x-ray images taken from two different angles as two different models.

Saik Anam Siam↗

Diagnosing plasma conditions in a copper wire array shot on Z : spatially-averaged analysis compared to inferred properties of individual bright spots.

Recent copper wire array shots on Z, when spectroscopically analyzed on a spatially-averaged basis, appear to have achieved ion densities near 10{sup 21} cm{sup -3}, electron temperatures of 1.25 keV, and K-shell radiating participation of 70-85% of the load mass. However, pinhole images of the shots reveal considerable structure, including several well-defined intensely radiating 'bright spots', which may be due to enhanced density, temperature, or some combination of the two. We have analyzed these individual spots on selected shots, using line-outs of their spectrum and inferred powers based on their images. We compare the properties of these spots (are they dense, hot, or both?), and examine their effect on inferring the radiating mass.

70 PLASMA PHYSICS AND FUSION TECHNOLOGY↗

Facility Activity Inference Using Radiation Networks

We consider the problem of inferring the operational status of a reactor facility using measurements from a radiation sensor network deployed around the facility’s ventilation off-gas stack. The intensity of stack emissions decays with distance, and the sensor counts or measurements are inherently random with parameters determined by the intensity at the sensor’s location. We utilize the measurements to estimate the intensity at the stack, and use it in a one-sided Sequential Probability Ratio Test (SPRT) to infer on/off status of the reactor. We demonstrate the superior performance of this method over conventional majority fusers and individual sensors using (i) test measurements from a network of 21 NaI detectors, and (ii) effluence measurements collected at the stack of a reactor facility. We also analytically establish the superior detection performance of the network over individual sensors with fixed and adaptive thresholds by utilizing the Poisson distribution of the counts. We quantify the performance improvements of the network detection over individual sensors using the packing number of the intensity space.

Rao, Nageswara S.↗

Multilevel Bayesian Inference for Cosmic Shear

The software implements a probabilistic hierarchical pipeline for inferring the gravitational lensing shears of galaxy images in the presence of noise, pixelated images, and astrophysical uncertainties. The distinct software components implement forward modeling of CCD images of galaxies and hierarchical inference of properties of the galaxy distribution and telescope Point Spread Functions (PSFs). We also include scripts to make diagnostic plots from the output files of each stage of the pipeline.

Bard, Deborah↗

Estimating Cosmological Constraints from Galaxy Cluster Abundance using Simulation-Based Inference

Inferring the values and uncertainties of cosmological parameters in a cosmology model is of paramount importance for modern cosmic observations. In this paper, we use the simulation-based inference (SBI) approach to estimate cosmological constraints from a simplified galaxy cluster observation analysis. Using data generated from the Quijote simulation suite and analytical models, we train a machine learning algorithm to learn the probability function between cosmological parameters and the possible galaxy cluster observables. The posterior distribution of the cosmological parameters at a given observation is then obtained by sampling the predictions from the trained algorithm. Our results show that the SBI method can successfully recover the truth values of the cosmological parameters within the 2σ limit for this simplified galaxy cluster analysis, and acquires similar posterior constraints obtained with a likelihood-based Markov Chain Monte Carlo method, the current state-of the-art method used in similar cosmological studies.

46 INSTRUMENTATION RELATED TO NUCLEAR SCIENCE AND ↗

Strong Lensing Parameter Estimation on Ground-Based Imaging Data Using Simulation-Based Inference

Current ground-based cosmological surveys, such as the Dark Energy Survey (DES), are predicted to discover thousands of galaxy-scale strong lenses, while future surveys, such as the Vera Rubin Observatory Legacy Survey of Space and Time (LSST) will increase that number by 1-2 orders of magnitude. The large number of strong lenses discoverable in future surveys will make strong lensing a highly competitive and complementary cosmic probe. To leverage the increased statistical power of the lenses that will be discovered through upcoming surveys, automated lens analysis techniques are necessary. We present two Simulation-Based Inference (SBI) approaches for lens parameter estimation of galaxy-galaxy lenses. We demonstrate the successful application of Neural Posterior Estimation (NPE) to automate the inference of a 12-parameter lens mass model for DES-like ground-based imaging data. We compare our NPE constraints to a Bayesian Neural Network (BNN) and find that it outperforms the BNN, producing posterior distributions that are for the most part both more accurate and more precise; in particular, several source-light model parameters are systematically biased in the BNN implementation.

46 INSTRUMENTATION RELATED TO NUCLEAR SCIENCE AND ↗

Count Every Trip: Finding the Uncertainty in Energy Estimates Made from Inferred Travel Modes

To properly inform transport policy and infrastructure changes, transportation related metrics need both measured values and uncertainties of those values. Travel monitoring smartphone apps can record people's travel behavior, but trip data quality is limited by sensor errors, user labeling rates and the accuracy of inference algorithms used for travel diary creation. We discuss the use of phone app recorded travel diary data to estimate energy consumption, and propose the use of propagation of variance to find error bars for such estimates. We define energy consumption for one trip as trip length times the energy intensity per distance unit of the travel mode used. We characterize trip length errors with relative error and inferred trip mode errors with confusion matrix columns. The resulting variances of each measurement are then propagated to the final calculated energy consumption. We tested our uncertainty methods on a dataset that used phone app data combined with prompted recall, consisting of 92,234 labeled trips for over 500,000 miles. Accounting for uncertainty using expected energy intensities and variance propagation gives a dataset-wide aggregate energy consumption percent error of about 9%, within one standard deviation from the truth. Future work could involve applying similar methods to other travel diary based metrics.

ADVANCED PROPULSION SYSTEMS,MATHEMATICS AND COMPUT↗

Count Every Trip: Finding the Uncertainty in Energy Estimates Made from Inferred Travel Modes

To properly inform transport policy and infrastructure changes, transportation related metrics need both measured values and uncertainties of those values. Travel monitoring smartphone apps can record people's travel behavior, but trip data quality is limited by sensor errors, user labeling rates and the accuracy of inference algorithms used for travel diary creation. We discuss the use of phone app recorded travel diary data to estimate energy consumption, and propose the use of propagation of variance to find error bars for such estimates. We define energy consumption for one trip as trip length times the energy intensity per distance unit of the travel mode used. We characterize trip length errors with relative error and inferred trip mode errors with confusion matrix columns. The resulting variances of each measurement are then propagated to the final calculated energy consumption. We tested our uncertainty methods on a dataset that used phone app data combined with prompted recall, consisting of 92,234 labeled trips for over 500,000 miles. Accounting for uncertainty using expected energy intensities and variance propagation gives a dataset-wide aggregate energy consumption percent error of about 9%, within one standard deviation from the truth. Future work could involve applying similar methods to other travel diary based metrics.

ADVANCED PROPULSION SYSTEMS↗