The Morning Report Video - transcript
The Morning Report is a proactive safety and system monitoring tool that applies a data-intensive computational approach to analyzing flight data for airlines.
Airlines are faced with many questions concerning safety and efficiency. Thousands of flights come and go every day with no way to monitor them all and make sense of the data...until now.
The Morning Report is a computational tool designed to analyze large sets of data collected by existing onboard instrumentation systems. It then performs the analysis overnight, and delivers the results each morning in an organized and prioritized report that highlights areas of possible safety issues.
With such comprehensive information provided in a readily usable form, aviation experts can better determine the answers to questions as they arise. Further, they can be inspired to formulate new, insightful questions as they search for ways to improve aviation safety based on real data. These insightful questions can help them find new ways to improve aviation safety.
The morning report is generated from gigabytes of data gathered from thousands of flights. The data are analyzed by software on a personal computer using complex, sophisticated statistical analysis techniques developed to analyze the data to identify flight characteristics. Most of the data show flights clustered together in typical patterns behaving in similar and expected ways during the phases of flight, such as taxi, takeoff, cruise, approach, and landing.
Some flights, however, cluster in less populous groups; just a few flights or occasionally, just one flight. Often these flights are not within the expected range of flight characteristics and are of special interest. The Morning Report highlights these flights to allow aviation experts to quickly interpret what happened on those flights to make them so unusual.
The Morning Report displays the data from these atypical flights in an intuitive format so that aviation experts can quickly understand and interpret the information. This new info often leads to new insights.
The Morning Report is issued every morning. The flight data are prioritized with the most unusal flights presented first.
The aviation expert can immediately see the difference between typical and atypical flight patterns within the intuitive graphical display for the phases of the flight. In this way, the display highlights the variables that are most likely to provide the insights as to what made that flight unusual.
An even better understanding of a particular flight can be gained when the user drills further into the data of any portion of any flight to view hundreds of other variables. These data are presented in tabular or graphical formats that can be expanded or condensed over time to pinpoint the exact moment or moments of interest.
The insightful data can be shared among aviation experts for collective decision-making or corrective actions
An exciting element of the MR is that this veratile data-intensive computational approach can also be extended to other domains that generate gigabytes of data. As in the aviation domains, the analysis of the data to identify unusual events can yield the insights needed to make specific decisions for improved outcomes.
In summary, the Morning Report uses sophisticated multivariate statistical analysis incorporated into user-friendly software. The user does not need to have a statistics background.
The Morning Report enables the user to understand typical patterns observed in the operation of thousands of flights and to pinpoint atypical events. These atypical events either can be envisioned anomalies, that is situations the aviation experts expected to monitor, or unenvisioned anomalies, situations highlighted in the data analysis that the experts may never have encountered or envisioned.
The insightful information derived from the patterns of flight characteristics may predict safety issues. Having this info allows the airlines to share it with the aviation community as appropriate enabling them to collectively formulate improved aviation policies and action plans that may prevent accidents and save lives.
The Morning Report was funded, supported, and guided by NASA Ames Research. PNNL/Battelle formultated the data-intensive computational approach and developed the algorithms for the multivariate analysis. ProWorks Corporation created the software incorporating the sophisticated multivariate analyses into an easy-to-use software with user-intuitive graphical displays.
Contact: Tom Ferryman, Pacific Northwest National Laboratory, 509-375-3888, tom.ferryman@pnl.gov