Aviation Safety Technology Portal (ASTEP)

These tools are not opensourced. For more information about these tools or to explore working with them, please contact the project owner on the tool's individual page.

ReADS ReADS can analyze text reports, such as aviation reports and problem or maintenance records. ReADS uses text clustering algorithms to group loosely related reports and documents, this reduces human error and fatigue. Plus, ReADS identifies interconnected reports; automating the discovery of possible recurring anomalies. ReADS provides a visualization of the clusters and recurring anomalies. ReADS has been integrated into a secure web-based search tool to allow uses to perform their own ... [read more]

Mariana is a text mining tool that can be applied to written reports such as ASRS reports, determining the likely categories that each report falls into (for example Runway incursion or Bird Strike), and calculating a confidence for each classification. Mariana’s innovation is that it automates the search for the best parameters to run the algorithm, instead of users needing to set the parameters themselves.

To learn more, please go here.

The SequenceMiner was developed to address the problem of detecting and describing anomalies in large data sets from recordings of switch sensors in the cockpits of commercial airliners. SequenceMiner works by performing unsupervised clustering (grouping) of similar sequences together, followed by a detailed analysis of outliers to detect anomalies. In this method, an outlier sequence is defined as a sequence that is far away from a cluster. SequenceMiner also includes new algorithms for outlier analysis that provide comprehensible indicators as to why a particular sequence was deemed to be an outlier. This provides analysts with a coherent description of the ... [read more]

Gaussian Process (GP) regression allows you to map the relationship between your input data and a target parameter. GP based tools can be used to predict failure or adverse events. For instance, with airplane trajectory data, a GP model could be used to predict a conflict of two flights paths. StableGP is a Gaussian Process tool that is highly scalable (can run faster on larger data sets), therefore making it more useful to the aviation community.

More information on this tool and how to run it can be found here.

Multivariate times series refer to time series data that has multiple variables. Surprisingly, research on MTS search is very limited. Most existing work only supports queries with the same length of data, or queries on a fixed set of variables. However, there are many multivariate time series data sets in aviation, for instance, a data base might include: flights of multivariate time series that show the settings of the control surfaces (usually discrete signals), the pilot inputs (discrete), as well as the heading, speed, and readings from the propulsion systems (all usually continuous). In many such databases, the number of ... [read more]

MPD ++ is a much faster version of Matching Pursuit Decomposition (MPD). MPD is a feature extraction tool. With some prior knowledge about the data, MPD can be used to extract some (specific) features of interest. The user can run MPD on the entire dataset to extract these features and then run detection/classification/regression algorithms on these features. MPD can be used as a dimensionality reduction algorithm to obtain a set of representative features that are relevant for the analysis. For example, if the goal is to find only the sinusoidal components with frequency ‘f ‘from a given signal, the user ... [read more]

Nu-Anomica is a One Class Support Vector Machine algorithm (OCSVM) that can be utilized on continuous (eg, engine temperature or fuel flow) data. A OCSVM takes a set of data, and for each new input in the data predicts which category the input might lie in (for instance: Anomaly, Not an Anomaly). However, many of the data sets involved in aviation are huge, and classical OCSVM’s can take a long time to run. Nu-Anomica has a much faster execution time compared to the classical version of OCSVM and yet produces a very similar outcome.

To utilize nu-Anomica, users will need ... [read more]

MKAD (Multiple Kernel Anomaly Detection) is a software tool that can be used for anomaly detection.

In most systems data may come in different forms, eg continuous, discrete, text etc. MKAD allows users to search for anomalies in data integrated from different sources and having different properties. For example, in Aviation research, MKAD might be applied to a series of data sets that includes continuous data like: engine temperature or fuel flow parameters, discrete data like: switches pressed by a pilot in an on/off sequence, and text data like: maintenance or pilots’ safety reports.

Examples of synthetic data files (with ... [read more]