Research Into Resilience: From Data to Decisions

Supply chains are too complex for spreadsheets. Tariffs shift weekly, geopolitical crises cascade through networks, and energy bottlenecks multiply. Organizations need systematic ways to map invisible dependencies and stress-test resilience before crisis hits.

That’s the research problem supply chain strategist Naman Shrivastava is investigating. Over the past year, he’s been building advanced modeling frameworks combining time-series forecasting, network optimization, and scenario-planning to surface structural vulnerabilities in global supply chains. It’s pure research—exploring what’s possible when you treat supply chain risk as a mapping problem.

“Can you systematically surface hidden dependencies that spreadsheets completely miss?” Shrivastava explains. “And if you can, what decisions does that visibility enable? That’s what’s driving this work.”

The Research Framework

Demand Forecasting & Inventory Optimization

The first component ingests demand data and applies multiple forecasting methodologies—ARIMA, Prophet, machine learning ensemble models—to generate probabilistic forecasts. The real research value is in optimization: calculating optimal inventory policies across networks while factoring in lead time variability and service requirements.

“The research asks: How much buffer is actually necessary at each node, given current volatility and geopolitical risk?” Shrivastava notes. “The modeling runs continuous optimization against real scenarios. What if lead times stretch 40%? What if tariffs spike on 15% of supply? The framework recalculates optimal inventory in real time.”

The models run hundreds of scenarios to test how resilience assumptions hold under different stress conditions.

Route Optimization & Network Design

The second domain maps how physical supply corridors interact with geopolitical risk, costs, and constraints. The framework ingests shipping lanes, transportation modes (air, ocean, rail, truck), and real-time corridor data, then models routing at scale across multiple modes simultaneously.

The key innovation is stress-testing. Models simulate what happens if the Suez Canal becomes congested, if Hormuz shipping costs spike 35%, if key ports face disruption. The framework recalculates optimal routes, showing which corridors become economically unviable and which alternatives emerge.

“The Hormuz crisis revealed companies had no way to answer: which suppliers route through which corridors, and what happens if specific corridors close?” Shrivastava explains. “The modeling maps that. It shows which suppliers ship through the same geopolitical chokepoints, which ‘backup’ suppliers aren’t actually backups because they use identical routing, where genuine redundancy exists versus where it’s just contractual.”

Supplier Performance & Multi-Tier Dependency Mapping

The third domain tackles supplier risk and multi-tier visibility. Models pull performance data—on-time %, quality, lead time variability, price—and segment suppliers across frameworks: Strategic, Bottleneck, Leverage, Non-critical.

The novel contribution is multi-tier mapping. Traditional supplier management sees one layer: your suppliers. This framework maps two, three, sometimes four tiers deep. If your supplier depends on a fab in Taiwan for components, and that fab depends on Qatari helium, and Qatar sources helium through Hormuz, the framework traces that entire chain.

“The research asks: Can you make invisible dependencies visible through systematic modeling?” Shrivastava notes. “Most companies know their direct suppliers. Very few know their suppliers’ supply chain risk. When you surface that, you see risks spreadsheets completely obscure.”

Why This Research Matters Now

Supply chain crises have stopped being anomalies. They’re structural features of the global economy. Hormuz tensions, permanent tariff volatility, energy scarcity, semiconductor shortages—these aren’t one-off events. They’re the baseline.

“The research asks: How do you systematically think about supply chain resilience when crisis is the baseline?” Shrivastava explains. “Traditional supply chain management was built for stable environments. When your environment is permanently volatile, the optimization problem changes completely.”

The research has generated unexpected findings: inventory is insurance, not waste, when lead times are long and uncertainty is high. Most companies’ backup suppliers aren’t actually backups—they share single points of failure. Geopolitical risk doesn’t surface in traditional procurement analytics.

Early Research Applications

Hyperscale Infrastructure Operators

Data centre operators face acute supply chain problems: grid connections slip 18–24 months, cooling vendors are booked 32 weeks out. The modeling frameworks map how these constraints interact, identifying which timelines are most at risk and where inventory buffers matter.

Semiconductor & Advanced Manufacturing

Fabs depend on specialty materials (helium, bromine), advanced equipment, and precise timing. Supply chains that mapped multi-tier dependencies discovered which product lines were affected by Hormuz crisis. Those that hadn’t found out reactively.

Procurement Under Tariff Volatility

As tariffs become permanent, procurement strategy shifts from “best price” to “optionality.” Modeling frameworks help procurement teams understand which supplier diversification actually reduces risk versus which adds cost.

What This Research Is Revealing

Organizations using these modeling frameworks develop fundamentally different mental models of their supply chains. They see vulnerabilities that spreadsheets miss. They understand where resilience matters. They make strategic bets on diversification based on actual chokepoint risk.

“The research is still in development,” Shrivastava emphasizes. “The goal is understanding: Can systematic modeling reveal hidden dependencies? What approaches work best? How actionable are insights?”

Whether this research becomes a commercialized tool is separate—one Shrivastava hasn’t focused on. The research itself is what matters: understanding supply chain resilience through advanced modeling, surfacing invisible dependencies, building frameworks for permanent volatility.

“If the modeling work generates useful research insights and helps organizations understand their supply chains better, that’s the win,” he says. “What happens after—whether it becomes a platform, whether it stays research—that’s not the focus right now.”

About This Researcher

This modeling work is conducted by Naman Shrivastava, a logistics and supply chain professional with an MSc in Logistics and Supply Chain Management from Sheffield Hallam University of England, an MBA in Marketing from JLU India, BBA in Foreign Trade from IIFTR. Naman also holds the honorary Membership of the Chartered Institute of Procurement & Supply also commonly known as MCIPS. His research combines forecasting, optimization, and network analysis to explore how advanced modeling surfaces hidden dependencies in global supply chains. Currently in active development and testing across infrastructure, semiconductor, and logistics operations.