Irish-born Whig politician Edmund Burke rightly said: “You can never plan the future by the past.” Investors know this only too well. That share price performance graph seems as if it will continue its upwards trajectory for ever. Then comes the crash. Those dreaded “unforeseen circumstances” scupper your rosy forecasts.

Scenario Planning, and the data science underpinning it, can help foresee many of these circumstances. Sophisticated software can work out the best sequence in which to perform each task in a project that can have thousands of steps spanning decades.

Scenario Planning generates a set of decisions over time based on a defined set of targets to produce an optimal plan. With the help of optimisation algorithms, we consider huge numbers of variables, from the macro, such as the political, regulatory and economic environment at the time, to the micro, such as the availability of diggers on a particular day, for example.

Many of the assumptions we make in creating these plans come with varying degrees of uncertainty. We can plan for these, too, running many scenarios each with unique combinations of all the values that carry some uncertainty. Then we can analyse how various aspects of the uncertainty impact the decisions we make.

But Adaptive Planning goes a step further.

Adapt to survive

While businesses can handle intrinsic uncertainties reasonably well – think of material and labour costs, for example – extrinsic uncertainties, such as the impact of climate change, war, or pandemics, are much harder to predict.

Adaptive Planning attempts to mitigate these ‘known unknowns’. We create plans that respond to the dynamic situations the world throws at us and build on our uncertainty planning.

Imagine several ‘what if’ scenarios ranging from the best to the worst, with lots of ‘pivot points’ along way: if this happens then we could do that, or something different, depending on the level of risk we are prepared to take and the amount of money we are prepared to spend.

In effect we create a ‘least regrets’ core plan that guards against future uncertainty. The pivot points enable us to adapt the plan as the future becomes clearer, while staying as close to the optimum path as possible.

Reservoirs are another example. They are expensive and time-consuming to build – the UK hasn’t built any new ones for 30 years – but what if a growing population and increased demand means water companies have to draw so much water from aquifers and rivers that biodiversity is badly damaged? Wouldn’t they need more reservoirs then? On the other hand, climate change seems to be leading to wetter conditions in the UK so perhaps the rivers and aquifers will cope. And water companies could redouble their efforts to plug leaks, encourage users to use water more efficiently, impose hosepipe bans and so on.

In other words, there are many scenarios that need to be planned for, often including thousands of variables over long time spans, decades potentially. This complexity is common to most large infrastructure projects, from offshore windfarms to fibre optic networks, from flood defences to nuclear power stations.

All could benefit from Adaptive Planning.

A plan that fails to adapt to new input data is asking for trouble. Yet many businesses, whether through short-termism, inexperience, or lack of time, are not able to consider all these scenarios. Which is why projects often go wrong.

Of course, there are also ‘unknown unknowns’, those unforeseen events described in Nassim Nicholas Taleb’s book, Black Swan, his metaphor for the unexpected, improbable event that has a profound impact. Such an event can be seemingly insignificant, but with the ‘butterfly effect’ can end up having a huge impact across a complex interlinked system over time.

Adaptive Resilience Planning attempts to cope with these black swan events, too, creating a plan that enables companies to ensure they have just enough ‘still left in the tank’ to respond to the unpredicted and recover quickly. This enables them to avoid plan fragility whilst fully quantifying and justifying the level of protection built into the plan.

The human touch

At WSP, we believe our expertise in Adaptive Planning can help companies plan multiple projects with more confidence, thereby reducing the risk that they will go badly off track. Ultimately, this saves time, money, reputation, and a lot of aggravation.

But it’s important to draw a distinction between Adaptive Planning and AI-powered predictive analytics. As AI is mostly trained on historic data it is as – if not more - susceptible to black swan events. Over-extrapolation from data that grows more unreliable the further we gaze into the future can lead to disastrous consequences.

Imaginative and emotion-driven humans are better than computers at identifying what goals we consider valuable, and what would constitute a worst-case scenario. Data science and AI can then help us achieve what we’re less proficient at, namely working out how best to achieve those goals.

By combining the best of data science, mathematics and AI, with the best of humanity, in an increasingly volatile world, Adaptive Planning, can help businesses adapt more nimbly to events and increase their chances of success and resilience into the future.