COVID-19 accélère la supply chain management analytics-driven

Des shortages d'équipements de protection personnelle à une large array de consumer goods, la pandémie COVID-19 a disrupted global supply chains in both expected et unforeseen ways. Experts estimate that a full recovery could take many months.

Pour rebondir plus rapidement, les supply-chain leaders tournent vers advanced technologies — IoT data, analytics, et machine learning — pour gain real-time insight into erratic supply et demand patterns.

« Having the right machine-learning and IA tools enables firms to understand market dynamics et manage their supply chains more effectively », a dit George Bailey, director of the Digital Supply Chain Institute.

La disruption began in China, the world's manufacturing powerhouse. By 2010, China had overtaken the United States in manufacturing output. While the SARS epidemic of 2002-2003 had China at 4.3 % of global GDP, MIT professor David Simchi-Levi notes it now accounts for 16 %.

Globalization As We Have Known It Is Over

Companies qui relied heavily on Chinese suppliers feel the impact acutely. Selon 2020 Statista data, some retailers source more than half their inventory from China. The same survey found that 44 % of retailers expect supply-chain delays, 40 % anticipate inventory shortages, et over half of electronics manufacturers foresee up to four weeks of disruption — an unacceptable lag in an era of two-day delivery expectations.

While many firms are scrambling to reassess their networks, risk management has historically been a low priority. The Institute for Supply Management's recent survey revealed that nearly three-quarters of companies reported disruptions, yet 44 % had no formal mitigation plan.

« The urgency to diversify et build redundancies is now a reality », a dit Alex Capri, senior fellow at NUS Business School. « The era of globalized supply chains as we once knew it is over. »

Bringing Analytics to Supply Chain Management

Manufacturers face cascading disruptions — from suppliers' plant closures et inventory shortages to transportation delays et workforce absences. Tier-1, tier-2, et tier-3 suppliers all experience setbacks, yet many firms lack visibility into these ripple effects.

« Because so many manufacturers depend on China — directly or through multiple tiers — factory shutdowns et understaffing have tightened capacity », a expliqué Bailey. A 2018 Statista survey indicated that 21 % of supply-chain professionals view visibility as a significant organizational challenge.

« Today most companies still rely on Excel to model scenarios », a dit Bailey. While Excel is useful, more sophisticated tools can deliver greater accuracy. A 2022 Supply Chain Quarterly study found that over 90 % of managers use Excel for analytics, whereas 82 % who employ advanced tools see marked improvement.

Deploying IA successfully requires ample data. « An IA system must be fed comprehensive datasets to learn how to respond », a noté Bailey. Ralf W. Seifert et Richard Markoff, dans « Demand for AI in Demand Planning, » argue that inconsistent forecasts across sales et operations can bias IA outcomes.

IA in Demand Planning

Bailey emphasized that demand planners need to grasp shifting demand patterns during crises et influence them. IoT sensor data, for instance, helps tire manufacturers monitor pressure et tread wear, proactively alerting customers et informing inventory needs.

« Technology et analytics are essential for understanding what drives demand et for IA to forecast future needs », a dit Bailey. While some firms must invest in new tech et talent, overall labor costs may decline as automation supplements planners' work.

Gartner predicts that by 2023, at least 50 % of global companies will deploy IA-driven transformations in supply-chain operations.

However, many organizations must first clean et standardize their data. « A lot of data is untrusted — misformatted, inaccurate, or biased — so considerable effort goes into making it usable », a averti Bailey.

Data quality remains the biggest hurdle for IA in demand planning, Seifert et Markoff observed. In the short to medium term, supply-chain managers will lean heavily on analytics to navigate scarcity et uncertainty.

Bailey suggested that companies may need to narrow their SKU portfolios, focusing on high-value customers et making fact-based prioritization decisions. « If you have 100 units of a product but need 500, you must decide which customers to serve first », il explained.

Strategic sourcing will also shift. « Overconcentration in China is no longer viable », a dit Bailey. Balancing risk et opportunity across geographic locations has become paramount.

Ultimately, the pandemic forces firms to adopt the technologies et practices they had postponed but now must embrace for supply-chain resilience.

« COVID-19 is a catalyst for change that many companies would have undertaken anyway », a conclu Bailey. « The good news is these changes are now inevitable. »