The framework

We work from a fairly simple premise: complex systems can usually be improved once you understand what state they are in, what variables matter, and how changes to the system affect the outcome. Sometimes those relationships are already well understood. Sometimes figuring them out is the project.

Our framework is a way of approaching both cases. At its simplest, it follows a recurring sequence:

SENSE → INTERPRET → AUGMENT

Sense

We start by defining the system and the question we are trying to answer. From there, we determine what information is available and what needs to be measured.

In some projects, the relevant signals are obvious. In others, we may need to develop a measurement, design an experiment, compare candidate signals, or establish whether a proposed measure actually captures the state or process we care about.

PNC_CORE provides the technical foundation for this work: acquisition, synchronization, data handling, computing and the shared tools that support the rest of the framework. PNC_SCAN applies those capabilities to multimodal measurement of human, machine, and environmental systems.

Interpret

Once useful data exist and are captured, the next problem is understanding what they mean. PNC_SCAN combines signals, tests relationships, develops models and determines which measurements actually help explain the system.

This can be a relatively direct analysis problem, or it can be a research program involving experimental design, repeated measurement, and validation. Psylinks is comfortable working in either case, and we can make progress even where the variables, models, or useful endpoints are still being established.

Augment

Once there is enough understanding to support an intervention, PNC_AUGS applies it. Depending on the problem, that could mean changing a workflow, providing information to an operator, adapting software, modifying a control system, or applying neurofeedback or neurostimulation.

The effect of that intervention becomes another measurement. We can then test whether it worked, refine the model, change what we measure, or change the intervention. That iterative process is what closes the loop.

Where this came from

Our first applications of this approach were in neuroscience. The brain and mind are difficult systems to study: they are dynamic, noisy, only partially observable, and many of the quantities we care about have to be inferred indirectly. Working in that environment requires good measurement, careful experimental design, signal processing, modelling, and a willingness to revise the question when the data demand it.

Those skills transfer readily to other complicated systems. Human performance, industrial processes, and energy infrastructure present different technical problems, but they can be approached through the same combination of measurement, experimentation, interpretation, and controlled intervention.

That is the capability PsyLinks brings internally to Olenox, and is the basis of our applied intelligence work for external partners. If you have a have a complex environment and need to turn its data into decisions and action, feel free to contact us at core@psylinks.ca.