how it works

Explore the Metacog platform, capabilities, and use cases.

 

 

The big picture

See an overview of the Metacog platform's innovative approach to capturing, organizing, modeling and analyzing training data.

Metacog Explainer Overview

 

graphic depiction with arrows and gear of IoT device data being captured

capture

Organized Secure Data Storage featuring cloud, keyhole and symbolic data

organize 

Model Deployment icon featuring model emerging from machine with gear inside

model

Analytic components icon featuring laptop with data in columns

analyze

 

The Metacog platform is made up of a set of APIs, a data lake, and machine learning components. Together, they are a toolkit for capturing new and valuable types of training data, organizing that data in ways that make it more meaningful, building models for monitoring and evaluating performance, and conducting analyses that lead to improvements in the performance of your people, teams and programs.

 

Schematic showing metacog platform including process data collection, storage, machine learning and analytics

 

Human Telemetry Data Components

Capture

Better data, smarter collection, and interoperability

  • Automatically captures, time stamps and fuses data from all your training hardware into a single time series. Mouse, joystick, VR headset, or biosensors like EEG? No problem.
  • Observes trainees as they work through a digital task, capturing not just their final answers, but also their behavior, process, and problem-solving approaches.
  • Records information about what trainee is seeing during training events (e.g., a plane entering the screen’s field of view).
  • Acts as a bridge between siloed training systems, whether they are used across teams or within a team at different stages of the training process.

hands holding game controller device near laptop keyboard that have been instrumented for data collection

Organized Secure Data Storage icon

Organize

Vast, secure, data lake—organized for insight

  • Session-centric approach to time-stamping and organization allows for data fusing from multiple devices, locations, and team members.
  • Uses smart, semantic encoding to ensure data has meaning (not "moved mouse" but "lowered flaps").
  • Built to handle from tens to tens of millions of simultaneous users.
  • Manages petabytes of data, processes millions of data point updates per second, and serves billions of queries that fetch trillions of results per day.
  • Retains a hypercube of sessions you can slice by user/team, by time, and by the sim software/scenario.
  • Provides playback of all session data and events, not just a video review.
blue and orange conceptual depiction of training session data stored in a hypercube format

Model Deployment icon

Model

SME-driven models offer unprecedented insight and scale

  • Allows your training experts (not data scientists) to define rubrics for good performance and input scored sessions into the modeling process.
  • Detects cognitive strategy and human factors; processing of data captured during training sessions gives a view into important affective states and behaviors (engagement, boredom, cheating).
  • Provides new visibility into hard-to-assess factors like situational awareness and cognitive load. (For more on situation awareness, see our SA whitepaper.)
  • Provides nuanced, real-time feedback to trainees and instructors, actionable during training sessions.
  • Scores sessions more quickly, accurately, and at a lower cost than humans, using SME-driven AI.
  • Offers innovative feature engineering, which allows faster, easier feature generation and winnowing by SMEs without the need for data scientists.
  • Unique scoring harness, IRR and DSM toolkit streamlines model management and helps avoid drift.
metal gears on a metal circuit board to convey machine learning modeling

Embeddable Analytic Components

Analyze

Powerful analytics embedded in your reporting platform

  • Provides immediate, accurate inferences for after-action reporting: get performance scores for individuals or teams plus deep diagnostics on why/how those outcomes were achieved.
  • Demonstrates impact of program elements such as content, tools, sequence, time, and other instructional factors on training outcomes.
  • Integrates Metacog's embedded analytics with your proprietary data for reporting in whatever reporting platforms you are using.
  • Enables descriptive statistics (what happened), diagnostic analytics (why it happened), predictive analytics (what could happen) and prescriptive analytics (what should be done next).
Male military customer accessing performance training analytics
 

use cases

Metacog is built for maximum flexibility and customization. Your needs determine what models are built and how they are tuned and deployed. Looking to assess situation awareness for fighter pilots? Improve sims for air traffic controller training? Evaluate readiness of elite warfighters? These scenarios would lead to different data capture, modeling and analyses activities, and Metacog's toolkit allows for all of them—and more. The use cases below are just a starting point. We would be glad to meet with you about how to apply the power of the Metacog platform to your training mission.

female military trainee

Individuals

  • Scale up accurate, rubric-driven scoring of performance tasks without having to add instructor time
  • Monitor and improve situation awareness
  • Detect cognitive load and adjust training demands accordingly
  • Identify trainee boredom and accelerate pace
  • Figure out which training behaviors are associated with successful outcomes

diverse four-person team benefiting from training analytics

Teams

  • Measure level of team collaboration (or lack of it) during multi-user activities
  • Identify and improve team resiliency in the face of stressors and obstacles
  • Assess whether hand-offs are occurring correctly in sequences where team members are responsible for different phases of performance tasks
  • Assess team readiness levels based on selected metrics

man with gear and process flows behind him to represent uses for training program improvement

Programs

  • Identify the most effective simulations for achieving specific training goals
  • Learn which "book learning" content is most associated with successful performance outcomes and what can be reduced or eliminated to increase efficiency
  • Determine whether changes in tools, locations, instructors, or other factors have a positive or negative impact on training outcomes
  • Predict impact of program changes on trainee readiness

Let us show you

We'll walk you through an example of how Metacog works using an air traffic control simulation tool and a Nervanix EEG headset. Watch in real time as Metacog captures the click-stream processes associated with managing takeoffs and landings and fuses them with sensor data captured from the headset.

 

GET A DEMO

screen shot of demo showing integration of EEG headset data with process data from air traffic controller training software
 

FREE WHITE PAPER

What if you could measure a trainee's situation awareness during a simulation session?

Situation awareness (SA) can be a critical factor in efficient, safe, and successful outcomes of high-stakes scenarios—from military missions, to disaster response, to operating rooms. Assessing and visualizing SA states is critical to effectively training personnel for operating in high-stakes contexts. But continuous, automated detection of SA in real time has been a challenge. The right kind of AI-based system can close the gap. Topics addressed in this white paper include:

  • What is situation awareness? 
  • Challenges in measuring situation awareness
  • The role of AI-based systems in measuring SA

download now

cover of Metacog Situation Awareness Whitepaper
 

Want to pop the hood?

Check out our demo widgets, documentation and tutorials.

show me

implementation flow

Metacog is a human-machine solution tailored to your specific training mission. We work with your subject matter experts to ensure smart encoding of data and the training and deployment of models that best support your needs. Talk to us about implementation

 

contact us