Simulation
Learn about how EDAA™ can enhance your digital twins simulations.
Simulation is an out-of-the-box functionality of Affective Computing by Virtue that lets you simulate an experience using Virtual Humans (VHs).
VHs are emotionally-driven non-playable characters powered by EDAA™ (our technology) that have psychological profiles based on those of real human users. Leveraging VHs enables you to replicate and gain insights from authentic emotional responses (similar to those of real humans).
Affective Computing by Virtue's simulation enhances your implementation of digital twins (the practice of designing models that accurately replicate prototypes of physical objects, architectural designs, models, processes, or services in the virtual world to study and predict their behavior).
Simulation helps you to analyze interactions between VHs and different aspects of the simulated experience.
Simulation is a big differentiator in the space of personalization, as using this feature, an experience and its human-centric outcomes can be validated even before it is made available in the real world.
How it works
First, EDAA™ interacts with real human users in real time when the experience provided by your solution is run.
Then, the experiences are simulated by running scripts. In the simulated, instead of real human users, EDAA™ interacts with Virtual Humans (VHs). VHs can also interact with each other during the simulation.
Each VH has a psychological profile that is either cloned or derived (through data augmentation) from a real human user's psychological profile.
Running the simulated experience includes invoking the personalization logics defined for each situation at the appropriate time. EDAA™ delivers the corresponding actions to the VHs.
After completing the simulation, you can gain insights and review the solution validation metrics by, for example, visualizing the results.
To learn more about Virtual Humans, see Entities.
Advantages
Simulation lets you:
Pre-validate a product or service that doesn’t exist yet. For example, you can validate the livability of a building, the walkability of a park, emergency evacuation routes based on an architectural model, and more. In these cases, as the final product doesn’t exist yet, you can use simulation to validate the experience's design.
Post-validate a product or service that already exists. For example, you can simulate a walkthrough of a digital showroom to determine the best image (that yields maximum engagement) to display at the entrance. In these cases, you can use simulation to validate and iteratively improve the experience's flow, content, and interactions with users.
Types of simulation
You can simulate experiences in the following ways:
Clone simulation
In clone simulation, each VH has a psychological profile derived from a specific real human end-user.
Clone simulation repeats the sequence of events observed in the provided data set. You can entirely clone an experience and use simulation to validate specific parts of it.
Bulk simulation
In bulk simulation, the VHs are not 1:1 clones of real human end-users; they are created based on data augmented from the psychological profiles of real users.
You can use bulk simulation to validate experiences that require thousands of participants, en masse.
How to provide initial (raw) data for the simulation
When implementing a simulation, you must provide EDAA™ with data about the experience. This enables it to understand which events it must observe during the simulation and what the boundaries of reality are.
You can do this in any of the following ways:
Regardless of how you provide your solution with initial data, EDAA™ augments the data during simulation.
Raw data collection (RDC)
RDC means running the solution experience with real human participants to collect the required initial data.
For example, if your solution is designed to validate the viability of evacuation routes in a public building by simulating humans' reactions to an emergency, you can orchestrate an RDC session in which:
There are at least 4 human participants (who represent the four main psychological profile archetypes)
Each participant wears a heart rate sensor
Each participant's verbal (speech) reactions are recorded throughout the experience
The emergency is emulated (for example, a fire alarm blares over a loudspeaker)
This would result in obtaining correct and unbiased raw data, as each user would react (verbally) and experience heart rate spikes in accordance with their psychological profile.
RDC is the most accurate and advisable method to collect initial data for simulation.
You must use a dedicated Affective Computing by Virtue project for RDC. During RDC, EDAA™ operates in passive mode (the project runs in user calibration mode) to collect data about human reactions to various situations. For additional information, see Importance of raw data collection for simulation.
By plugging in an existing data set
You can do this if you need to simulate an experience for which relevant data already exists.
For example, when validating how friendly a city is towards bicycle riding, you can use the large volumes of freely-available public data related to bicycling.
However, plugging in an existing data set introduces the disadvantage of EDAA™ never diagnosing uses (as the data belongs to riders who haven't used your solution).
By using completely synthetic data
You can generate a synthetic data set and plug it into your solution. However, you should choose this approach only for fast, low-level experience validation where delivering accurate personalization is not a priority.
We do not recommend choosing this option as it is inaccurate from the perspective of human centricity. Ideally, all the data used as the foundation for simulating a human experience should be from and about humans, starting from heart rate patterns, for example.
Last updated
Was this helpful?