GlobalABM technology 

Agent-based simulation creates a full-scale virtual world that explains the behaviour of a system by simulating the behaviour of each individual entity; be it a person, object or process.

In this way a user can better understand their sector and gain a competitive advantage through powerful insight into the future effects of today’s decisions.

Ease of use

  • Cloud based – no expensive software to configure and maintain

  • Simulator runs fully customisable - You are in control

  • Plug and play

  • Use market/industry data or your own proprietary data

  • Use our behavioural and decision making logic or select your own

  • Connect to your other existing systems via APIs

Key criteria for analytics

  1. Expressiveness: support for complex decision-making capabilities and interaction dynamics

  2. Usability: support for easy manipulation of models with the flexibility to achieve these quickly and without requiring extensive support 

  3. Performance: support for large-scale simulations (millions of agents)

Smart analytics

  • Why/when do prices go up or down

  • Identify potential volatile inflexion points

  • Identify supply & demand constraints

  • Identify logistics bottlenecks & congestion

  • What do these mean for prices and supply chains

  • What do they mean for prices of goods and transportation

  • What can you do to mitigate against adverse events

  • What might happen today; tomorrow; next week; next month; next quarter;next year ; next decade?

Dynamic analytics 

  • Simulate reasoning in a causal way

  • Drill down to examine the minutia of info

  • Carrying out tens of millions of calculations and iterations every second

  • Continually running and evaluating – the end goal is not a static equilibrium.  Whilst neat and tidy...the real world is rarely in this state

  • A continual self-reinforcing loop of simulated predictions > decisions > effects

    • Never stationary

    • Always evolving

    • Forever informing

Test & evaluate scenarios

Explore all possible scenarios and see how they are likely to play out across all the many complex inter-related dynamics that exist in the real world.  Foresight is knowledge; knowledge is power and speed of decision making.

Examine:

  • Outcomes of your decisions

  • Outcomes of other’s decisions

  • Interaction between these

  • Knock-on effects and iterative results and interactions

You are in control

  • GlobalABM – Your simulator that you control; built using our intelligent components.  We do the hard work, you reap the benefits

  • Only you have access to your simulator, its outputs and your proprietary data

  • Combination of our world beating technology and your knowledge and expertise gives powerful super-human results

  • Gives you the freedom to focus on decision making rather than analysing big volumes of data

The Power Of Simulation

Our proprietary technology

Agent-Based Simulation, Causal Analysis, Machine Learning, Neural Networks, Game Theory, Decision Trees, Optimisation and Heuristics all have their individual strengths and weaknesses.  Rather than exclusively use one technique, we have developed our own proprietary technology utilising the best elements of each and applying them for the tasks they were primarily intended.  

 

Our approach embraces complexity

 

A simplistic modelling approach can be OK for some tasks, but its limitations must be understood and recognised. 

 

Where insight is sought into complex real world dynamic, constantly evolving actions, reactions and interactions, a more sophisticated approach is required.

 

The world comprises many individually complex systems, sectors and societies that are computationally irreducible.  That is to say, they cannot be accurately described mathematically in a model.  COVID; classical economic; and energy market modelling are but three highly topical examples.  

 

Many people make the mistake of thinking everything can be described mathematically or compressed into simple formulaic shortcuts and thus use this approach to try to solve every problem, including complex mathematically irreducible ones.  This is wrong and usually costly, resulting in incorrect forecasts and insights.  As convenient and comfortable as this approach is, you cannot accurately model complexity through such simplified solutions.

 

For market leading accuracy, you need to ditch these incorrectly applied methodologies and embrace simulation.  Only then can you capture all the complex 3-dimensional step-by-step actions, reactions and interactions as they exist in the real world.  Only then will you see the accurate evolutions and emerging patterns that are vital to making present decisions that impact future revenues and opportunities.

 

The following are some limitations of traditional modelling and analytics that our ground-breaking approach resolves:

 

  • Treating key inputs either as being homogenous units (every person or every asset is the same)

  • Having defined linear relationships (If A goes up by 10% then B will go up by 5%, or e.g. assuming all ships will each make the same number of journeys each year)

  • Discrete, linear relationships (A follows B, follows C, follows D)

  • Statistical extrapolation of past data (I recognise this pattern from last year so I assume the same will happen again this time round)

  • Seeking static equilibrium end games (stable prices, supply=demand, perfect capacity utilisation)

In our computer simulated virtual world we fully incorporate all the dynamic complexity that exists in the real world.  We simulate all the actions, reactions, interactions between various entities and we never assume a static equilibrium is the end goal.  All entities that can make decisions; have independent behaviours and characteristics; specified limitations or restrictions are simulated by heterogenous, (individually unique and accurate) agents. We then incorporate the where/why/how of their decision making.  This is then all layered on top of a precise digital twin of the real world.  For example, if we are looking at the oil markets, we would include agents to represent all oil fields, ports, sailing routes, oil tankers, refineries and so on to build an exact virtual world replica of the real world. This is also referred to as a digital twin.

 

This allows the user to do two critical things

 

Firstly, they can simulate what is likely to happen as things currently stand, either based on data sets we provide or by incorporating their own proprietary data in a secure manner (only the user has access to this data and the simulation outputs generated).  The user is always in control.  

 

Secondly, users can wargame that which may happen by setting the simulator up to run any number of possible scenarios.  The possibilities are only limited by your own imagination.  For example, do you want to see what all the resulting effects of a major shock are? (e.g. closing the Suez canal, a major weather event, COVID impact).  You can do that in minutes with our simulator.   You can see what is likely to happen in the next few days and any time horizon thereafter that you want to explore.  Not only are the initial – and perhaps obvious – reactions simulated, our approach also lets you immediately assess the 2nd, 3rd, 4th,…, nth impacts and events – they are normally less obvious or even impossible to quickly assess.

The combination of these two critical factors enables superior actionable insight so that users can proactively identify opportunities and/or risks for mitigation. 

 

  • Identify inflection points in the market & capitalise from this by being the first to move/react

  • Have Actionable insight – We augment human decision making rather than replace it

  • Execute smarter and quicker decisions than your competition, be one step ahead – because you know the outcome of your actions and the actions/reactions of the market  

  • Discover and quantify trading opportunities such as hitherto unseen bottlenecks or supply chain issues that might manifest themselves and/or whether prices are likely to change away from current forward market assessments

  • Identify, monitor & plan to mitigate and/or capitalise from market/business risks

  • Monetise your proprietary data, & superior domain knowledge to become the market leader

Our technology is based on continuous interactions between agents and the landscape (digital twin) they operate in.  As such, our approach is not limited by first needing to build a new data set after a shock or structural change has happened before any attempt at forecasting can accurately be made.  This is a limitation of traditional forecasting tools.  Additionally, as we are not data reliant, we are not as exposed to issues such as missing or erroneous data.  The well-known “Garbage in – Garbage out” phenomenon is not one that our technology suffers from given that we focus on how agents react to events rather than extrapolating future forecasts based on past data. 

 

If you want to understand more about how our innovative and powerful technology can be applied to your markets, supply chains, ship/cargo movements to name but a few suited applications get in touch now for a no obligation discussion with our experts.