Frequently data values for inputs are within multiple, disparate, and often
isolated sources. For example, within seconds following onset of a bad
situation, an electric-grid operator might need to see data from systems
unconnected to each other – such as voltage readings and settings from
the control system, updates from the outage management system,
histories from the maintenance system, and forecasts from a
weather service.
For outputs there can be many to choose from and their relationships to
goals are not always obvious.
By applying AI technologies in real time to fuse inputs, outputs (in the form of Key Performance Indicators, or KPIs, linked to goals), visualizations (like maps and charts), and query tools, the result is a valuable situational awareness solution.
2. Situational Intelligence
This takes the data being “seen” via awareness to a “connecting of the
data dots” for intelligence.
With intelligence, SmartCloud applies AI reasoning techniques to cut the
time it takes operators to identify, predict, diagnose, and respond to
emerging situations. Humans can receive explanations of what situation
is occurring and why. Reasoning techniques might include rules,
semantic reasoning, machine-learning and other advanced methods
like Bayesian probability networks or simulation.
3. Situational Control
This level determines and takes actions that achieve optimal operational
and financial goals. It extends awareness and intelligence. It applies AI
reasoning and algorithmic techniques together with workflow
orchestration. Techniques might include AI Agents, optimizers, or
genetic algorithms.
Once a Situational Control solution determines the optimal course of
action, it can prescribe the steps to take as recommendations to
operators or it can automate the steps.
What's your situation?
At SmartCloud we’ve found that a situation-centric paradigm for
managing time-sensitive operations to be extremely powerful.
It’s what humans intuitively do.
Our AI-driven solutions supercharge the powers of humans to
identify and respond to situations of interest within business and
industrial operations. Our AI is designed from the ground up for
analysis of complex situations in real time.
So the question becomes “What are your situations of interest?” This
is what we ask first before starting a new solution. We’d welcome a
discussion about your situations and goals so that together we can
consider the potential of AI-driven situational management.
When the game’s on the line...
Situational management is critical
Here’s a tense situation, familiar to many U.S. football fans. It’s fourth
down, just inches from the goal line, and just seconds left in the game
as the clock ticks down. The team is trailing and needs to score a
touchdown to win...
The play is sent in from the coaches, and fans are holding their breath.
On the snap of the ball, the running back takes it from the quarterback
and leaps high in the air, toward the goal line ... and scores.
For the winning team, this is situational management at its best.
The coaches were well aware of the score, the time left, the distance to
score, the abilities of their players at that moment matched up against the
opponent’s players, the opponent’s history for similar situations, the
condition of the grass, the weather, and many other factors relevant to the
situation at hand.
The coaches had the intelligence to reason with all these factors, in real
time, to predict what might happen for alternative plays...
And they took control by concluding which play gave the best probability of scoring and then prescribing that play to the players, who then successfully executed it as a tightly orchestrated set of body-bruising actions.
Examples of the importance of well-done situational management abound. Beyond U.S. football, it is essential to success in soccer, hockey, basketball, volleyball, and many other sports. In business, it is vital to preventing fraud. In our everyday lives, we manage driving situations in real time as we seek to arrive at destinations safely, quickly and economically.
In Industrial Internet of Things (IoT) systems – like an electric grid – situational management is key to ensuring reliability and efficiency.
At SmartCloud, we are leaders in applying artificial intelligence to situational management. We create solutions for mission-critical operations in business and industry that improve real-time management of complex situations.
To better understand the value of what we do, let’s define what a situation is and introduce the role of AI.
It's a situation
At SmartCloud, we consider a situation to be a pattern, or combination, of inputs and outputs of interest within a managed system.
What makes a pattern to be of interest includes:
-
Alignment with goals -- such as operational performance, costs, or safety
-
Multiple states -- can be a problem, opportunity, or normal situation
-
Changes over time -- has a past, present, and future
-
Quantifiable -- the inputs and outputs can be quantified in a useful way
The managed system is what’s being controlled to achieve goals. Situational management is the making of changes to that system in order to reach desired situational states and associated goals.
Your car's situation
To illustrate these concepts, consider your car as the managed system.
Your goals as the driver are to arrive safely, on time, and at minimal
fuel costs. Situations of interest might include: unsafe speed during
bad weather, high risk for a speeding ticket, excessive fuel consumption,
or being late to your destination due to heavy and slow traffic. Hopefully
the situational state is that all is normal – you’re safe, ticket free and
on schedule.
Situational inputs are numerous – accelerator position, gear, road types, road maps, speed limits, other cars, weather conditions, and so on. Outputs include car speed, engine RPMs, fuel consumption, etc.
Situational management in this example involves making decisions and taking actions for your car – how far to push down on the accelerator at any moment, which gear to use, which route to take, and so on. Do it well and you’ll get to where you’re going safely, on time and with the most fuel left in your tank. Do it poorly and you might have an expensive traffic ticket, find yourself in a ditch, or be late.
Be Aware, Be Intelligent and Take Control
Based on decades of experience in building AI-driven solutions for some of the world’s most complex operations, including those within utilities, governments, telecommunications, aerospace, manufacturing, oil and gas, and finance, we at SmartCloud have come to realize that successful situational management requires three levels of reasoning:
1. Situational Awareness
This is about “seeing” situations and their many inputs and outputs.
When operators can fully see, they can more effectively apply intelligence
to and control situations of interest.
However, visualizing the data for inputs and outputs in a meaningful way can be very difficult.
December 3, 2015










© SmartCloud, Inc., All rights reserved.