Finding Needles, Understanding Uncertainty and Building Puzzles
We use a variety of analysis tools to extract new information from large data sets.
Finding the Needle in the Haystack
Data capture from interconnected measuring devices can now generate unmanageable quantities of data each day. Somewhere in that mass is information describing trends, faults or errors that could indicate impending failures or best operating practices. Finding those indications in a timely manner can enable corrective actions that could prevent failures or improve efficiency.
Understanding Uncertainty to Increase Value
Every piece of data has some degree of uncertainty due to factors such as measurement error, discrete sampling of continuously changing processes and misinterpretation. Probabilistic analysis methods can be used to account for a variety of uncertainties. This analysis provides a statistics-based description of the process that can be used to make informed decisions.
Many data sets are an assembly of discrete pieces of information that describe a set of assets. This information is often gathered from different sources and is in various formats. There may be no consistent structure or meta-data that links one piece of information to a specific asset. Statistical methods are used to analyse each record in the data set to determine what aspect of the asset it describes and which asset it applies to. This provides information supports operational decisions and enables performance tracking.
Various data analytics techniques are used in each of these scenarios including:
- Informed Searches
- Baysian Alalysis
- Monte Carlo Simulations
- Internet of Everything/Big Data
- Cloud Computing