New Analytics For A New Class of Data
Explosion of streaming position and movement data, with dynamic time-stamped attributes makes traditional approaches obsolete.
Leaders of industry know to be competitive, they must look forward (not backward). They must identify and
understand opportunities right now, and spend less time looking in the
As an emerging industry category, Position and Movement Analytics
describes the enabling technologies and business processes that
facilitate effective action from the statistical analysis of real-time,
ongoing, location-specific data feeds from sensing devices. These
devices include smart phones and devices, and state sensors which stream
Position Facts that include a wide variety of data in addition to
position and movement characteristics such as temperature, pressure,
humidity and fluid level, as well as still and video imagery. Position
Fact data is transmitted and received in real-time via a range of
wireless methods and network carriers.
The following answers to common questions outline the fundamental need for re-thinking analytics in general, and state the case for a new real-time opportunity focus of analytics instead of the historical and forensic business analysis so common today.
What are Position Facts?
Facts are time-stamped data packets that include information such as
movement characteristics (e.g., location, momentum, trajectory,
acceleration, speed and so on), and state/condition data specific to the
sensor (e.g., temperature, pressure, humidity, container seal/breach
status and so on). The exponential proliferation of smart phones,
appliances and sensor devices has created a torrent of rich new data of a
unique, ephemeral class. The richness of the data’s content is based
on both position/location and on reported “facts” correlated to context.
Value is defined by how recent the data is, and it may rapidly lose
value over time at a rate relative to the industry it may be serving.
The collective historical Position Fact data set provides another
untapped asset for some industries.
What can I do with Position Facts?
Facts, as a new class of data, enable the creation of ongoing real-time
position and movement signatures for a specific object (asset, resource
or person). A signature is a personalized statistic that captures the
object’s location and movement patterns and changes slowly over time. A
well-formulated signature captures the typical behavior of the object in
terms of the state and condition and other attributes in the Position
Facts. Signatures are adaptive, are updated continuously and
stochastically, and must self-initialize. As it is with real-time
streaming data, it is not possible for any manual intervention. Further,
a well-formulated signature built with the VisTracks platform can be
extensible. An extensible signature can grow with the object’s evolution
over time. In other words, as the object’s Position Facts are extended
with new additional attributes, its individual signature can be
Signatures can be analyzed, and in aggregate, create known (standard)
patterns, which can be used to categorize solution-specific position and
movement behavior. Many patterns can be collected into a profile
library unique to the interests of a specific solution. Profile-based
methods for signature processing can identify a particular behavior
pattern against a profile library. Then, as each signature is updated
with real-time Position Fact data, it is compared to the profiles to
determine if there is a match. For change detection, the signature is
compared against itself, or a computed set of the most recent Position
Based on these and other types of operations on signatures, statistics
can be computed on a population, correlated and analyzed, and then
generate appropriate notification to take automated action.
Why is Position and Movement Analytics better than historical analytics?
moves in real-time. Today opportunity data takes many forms, comes
from many sources and can effect a change in direction in real-time.
Traditional analytics software focuses on historical ERP, CRM and other
out-of-date business process data, typically from last year. By the time
analysis is completed, decisions based on that analysis are even more
out of date. Position and Movement Analytics enables effective
decision-making and action-taking based on correlating current,
location-specific time-stamped data. Native to the cloud, and
capitalizing the scalable statistical algorithms real-time digital feeds
are the intelligence input source for predictions as trends happen,
where they happen.
Why are traditional GIS-centric approaches obsolete?
GIS systems consider the map the center of the system, but for
analytics, it's the position and movement data that must be analyzed,
not the map. GIS, at best, only serves to provide the presentation
canvas upon which to visualize dynamic analytics with a location
component. Current systems are capable of putting push-pins for
locations on the map, creating a static representation, but for dynamic
real-time analytics, the data must be visualized in real-time too,
making static maps no longer relevant, the moment they are created.
Predictive analytics based on position and movement of an asset,
resource or people, is inherently dynamic, and the demand for tools that
leverage a statistical engine to compute on the location information in
real-time are key to capitalizing on future opportunity, as it happens.