How dogs catch frisbees – how to USE digital-twin models

International Aid programs

I recall work for a senior Aid-program leader at the UK’s Dept for International Development (now part of the Foreign Office), showing how dynamic models could radically improve the planning and delivery of development Aid programs.

This ended with a mini-conference to review this and other ‘systems’ approaches to managing Aid programs.

Now appreciate that this body, along with USAID and other foreign Aid bodies, is dominated by economists – so they wheeled out 2 top economists to review the results.

How dogs catch frisbees

I still recall this guy saying “These models can’t work – the future will always turn out different. Think how a dog goes about catching a frisbee. It constantly watches where the frisbee is going and adjusts its track to intercept it.” He’s not wrong on that part – see for example https://pubmed.ncbi.nlm.nih.gov/15200626/.

… but he could not have described BETTER how a digital-twin model can and should be used.

How digital-twin models (for business, Aid programs or anything) can and should be used

• Continually updated with emerging information, the model constantly adjusts the program’s likely trajectory (or the trajectory of business outcomes).
• Program managers can then adjust spending and effort – e.g. exploiting things that seem to be going better than expected, or changing plans and resources for things not going so well.
• But a digital twin model can do something more that the dog can’t – it can explore, ex-ante, plausible X-factors (like wind hitting the frisbee from behind a tree, or, say, the loss of a key partner in the Aid program).

How are Aid programs actually planned?

… by deploying the antiquated and hopeless “logical framework” or “LogFrame” (rather than ping any particular promotion of this method, search the phrase yourself) This simply assumes Inputs of money and effort lead to Outputs that deliver beneficial Outcomes – a dumb dog that just sets out for where it first sees the frisbee and keeps going, for 3-5 years!

Of course, smart Aid-program leaders do recognise the limitations of this static, linear planning method and do try to adjust programs as events unfold – but the program plan originally set is really hard to renegotiate, especially with no evolving, digital-twin view of the future to make the case.