Internet of Things (IoT) data and machine learning may overcome the failings of traditional supply chain risk management systems, which are falling short in a changing world, according to US analytics company Transvoyant.

“By continuously collecting massive real-time data streams from IoT devices around the world, these solutions enable risk managers to be nearly everywhere at once,” the group says in a white paper.

Real-time information enables managers to track every stage of the supply chain, and shows risks from natural disasters, strikes, civil unrest and other events.

“These solutions enable organisations to evaluate both internal and external risks by priority, as scored by advanced machine learning algorithms,” it says.

Global supply chain risks are at historically high levels, particularly with manufacturers in developed countries moving production offshore to low-cost regions.

But a 2014 survey by the Global Supply Chain Institute found 90% of 150 businesses surveyed did not quantify risk when outsourcing production.

Transvoyant says disruption costs on average $US73 million ($95 million) a year for 69% of Fortune 500 companies, while the Business Continuity Institute found 70% of global executives surveyed last year reported at least one disruption in the previous year. Of those, 43% were not insured.

The white paper, on the Risk and Insurance Management Society website, says supply chain risk approaches using new technologies offer benefits beyond avoiding disruption.

“It’s about continuously predicting the future across a living, breathing, complex and interdependent supply chain to achieve superior performance and gain competitive advantage,” it says.