How Merra works

Six principles, one loop

Whatever the technology underneath, every Merra system follows the same six principles: in small-scale fisheries, aquaculture, conservation and management, or vessel simulation and training. Each turn of the loop makes the next prediction better.

The Merra loop

A loop, not a pipeline

We develop technology, processes and systems that combine modelling and simulation with validation through cost effective sensors and remote sensing.

Text for the Merra loop above. Edit the wording here and press Save. This section is hidden on the live site. Keep six principles, each with a heading, a short line, a description and four examples in the same order as the sectors.

Sectors

  • Small-scale fisheries
  • Aquaculture
  • Conservation & management
  • Vessel simulation & training

Sense

Low-cost sensors and easily accessible apps collect real-world data, automatically wherever possible.

Our own sensors and field apps gather data at sea and on the shore, alongside open data from satellites, weather and ocean models.

  • Trackers and a catch app record where vessels fish and what they catch and retain, giving highly spatially resolved fishing effort.
  • Loggers, sampling apps and on-farm sensors record conditions, growth and stock biomass.
  • Tags and field apps record animal movements, strandings and sightings.
  • IMU and GPS sensors record how real vessels move at sea.

Simulate

Model how the system works.

Models built on biology and physics let people explore how a system behaves and what might happen next.

  • End to end stock assessment models of crab and lobster populations.
  • Digital twins of shellfish and finfish farms, nested in hydrodynamic models.
  • Computational Fluid Dynamics tests the fine scale hydrodynamic response such as drag caused by our biologging tags.
  • Vessel avatars built from real designs, in controllable wind and waves.

Validate

Models and simulations are tested against real-world data and their error is reported.

Results are compared with independent observations, from the field and from space, to check they hold up.

  • Assessments are compared with measured catches, recorded landings and buyers’ notes.
  • The twin is checked against each new sample and logger record.
  • The swimming avatar is checked against the tag’s own record.
  • Motion recorded on real vessels provides the test for the simulation.

Predict

Forecast, with honest uncertainty.

Tested models look ahead, showing the likely outcome and the uncertainty around it.

  • Stock status, and the likely effect of management options.
  • Line and farm level production on mussel farms.
  • How tag shape and placement change drag.
  • How a vessel is likely to respond to a given sea state or loading configuration.

Inform

Turn data into useful information for those that need to use it to make decisions.

Results are presented as clear visuals, dashboards and briefings for the people who use them.

  • Portal views for skippers, fishery officers and policy makers.
  • Farm and cooperative dashboards.
  • 3D avatars, driven by the tag’s IMU data, that show how the animal actually moves and behaves.
  • An interactive simulator for testing and training.

Decide & monitor

More informed decisions linked to ongoing monitoring.

People take the decisions; Merra helps track what happens afterwards and feeds it back into the loop.

  • Management measures, and their effect on catches.
  • Harvests planned, then logged.
  • Tag design refined for the next deployment.
  • Scenarios rehearsed in the simulator, and real trips recorded to understand real world responses.

Principle 1

Sense

Low-cost sensors and simple apps collect real-world data, automatically wherever possible.

What we do

We design, prototype and build our own low-cost sensing solutions and field applications. By drawing upon a range of open source data to augment locally sensed data we can model and simulate complex systems. By automating many processes and systems we can provide rapid, cost effective results, and information to support decision makers.

In practice · Small-scale fisheries

Merra-track, hub and portal together with Merra-catch, scan and log provide a comprehensive fisheries monitoring platform.

Merra toolsMerra-trackMerra-catchMerra-scanMerra-camMerra-logMerra-tagMussel SamplerStrand

Alsofarm loggers · satellite images · animal tags

Principle 2

Simulate

Model how the system works.

What we do

Combining biological and physical processes based on best available scientific knowledge and available data allows us to model and simulate complex systems. Our systems offer guided and simplified tools allowing non-specialists to understand and explore operational scenarios and make intelligent forecasts.

In practice · Wildlife

Designs and prototypes can be tested in virtual Computational Fluid Dynamic (CFD) environments to allow better understanding of the way fluids interact with structures and the forces involved.

Merra toolsMerra-driftMerra-shellfishMerra Stock DeskMerra-3D

Alsooil drift · farm growth · crab and lobster stocks

Principle 3

Validate

Models and simulations are tested against real-world data they were not built from, and their error is reported.

What we do

Ensuring that a model or simulation is a reasonable reflection of the real world requires independent validation. This is achieved by comparing models and simulations with real world data, observed or directly sensed as well as remotely sensed via satellites for example.

In practice · Oil and pollution

Merra-drift uses particle tracking modelling to forecast or hindcast where particles, objects or substances travel in the ocean. Satellite imagery can be used as real world evidence against which to test the validity of the model and ultimately to understand and where possible reduce uncertainty and correct errors.

AICalibration tunes model settings to match the observations.

Merra toolsMerra-logMerra-hubMerra-driftMussel Sampler

Alsofarm samples · measured catches · tag records

Principle 4

Predict

Forecast, with honest uncertainty.

What we do

Once a model has been tested it can look ahead. Every Merra forecast comes as a range, together with the data and assumptions behind it, so people can judge for themselves how far to rely on it.

In practice · Shellfish farms

Merra-shellfish creates a digital twin of a mussel farm site to simulate the growth and production of mussels. Production forecasts are corrected by periodic sampling of mussel rope dimensions and measurement of mussels.

Merra toolsMerra-shellfishMerra-driftMerra Stock Desk

Alsowhere oil will beach · stock status · tag drag

Principle 5

Inform

Make it clear to the people who decide.

What we do

Analysed results are translated into visualisations, dashboards, briefings and reports. Outputs are designed to meet the needs of those that need this information to be accessible and understandable to non-specialists.

In practice · Small-scale fisheries

Merra-portal shows individual fishers information relevant to the sustainability and profitability of their business. Regulators have access to information that can inform management and compliance.

AIPlain-language summaries and situation reports are drafted automatically, then checked by people.

Merra toolsMerra-portalFarm and cooperative dashboardsSituation reports3D avatars

Alsofarm dashboards · incident reports · tag avatars

Principle 6

Decide & monitor

People act, then watch what happens.

What we do

Merra processes and systems are designed to inform decision makers and to support post-decision monitoring and evaluation.

In practice · Shellfish farms

Merra-shellfish is designed to predict the production of mussels and when to harvest them. Buyers then record the weight of mussels actually supplied by the farm. The difference can be analysed to correct the growth model for a given site.

Merra toolsMerra-logMerra-trackSatellite monitoring

Alsomanagement measures · incident response · tag design

Put the loop to work

Tell us about your fishery, farm, coastline or study, and we will show you where Merra fits.