Armada has completed three AI-focused capstone projects with graduate students at Carnegie Mellon University's Heinz College of Information Systems and Public Policy, targeting food safety compliance, inventory optimization, and delivery accuracy. The collaboration, announced August 24, 2026, pairs Armada's operational supply chain data with Carnegie Mellon's academic research capacity.
The projects gave students experience with real-world logistics problems while helping Armada develop solutions that could be deployed across its operations. Armada provides supply chain services spanning freight management, inventory optimization, network analytics, and real-time visibility.
Food safety compliance
For food safety and traceability, the team developed a hybrid AI system to classify products under requirements including FDA FSMA 204, USDA regulations, and California Proposition 12. Armada said the system achieved 90% automated classification accuracy for items requiring advanced review.
The system reduced manual analysis by 70% to 85% and cut processing times from days to minutes. That matters for operations teams facing growing regulatory complexity across federal and state requirements.
Inventory and delivery accuracy
A second project used AI-driven modeling to optimize safety stock across restaurant supply chains, balancing product availability against excess inventory and food waste. The third initiative focused on estimated time of arrival accuracy, using shipment and GPS data to build a machine learning model that reduced mean absolute error by 53% - an average improvement of 124 minutes across 1.76 million predictions.
Those results point to practical applications for AI for Operations teams managing perishable goods and customer expectations around delivery windows.
What Armada says
"These projects reflect how Armada is thoughtfully embedding AI into the core of our business," CEO Chris O'Brien said. "We are applying AI and advanced analytics to tackle complex supply chain challenges in ways that are practical, measurable, and built to scale."
Armada CIO Katherine Karolick said the Carnegie Mellon collaboration helps the company accelerate development of AI-based capabilities while exploring new approaches to supply chain decision-making. Operations leaders looking to build similar capabilities can follow structured approaches like the AI Learning Path for Supply Chain Managers to develop the skills needed for these projects.
Why this matters for operations professionals
The measurable outcomes - 90% classification accuracy, 70-85% less manual analysis, and a 53% error reduction in delivery predictions - offer benchmarks for what AI can deliver in supply chain settings. The projects also show a working model for collaboration: companies with operational data and academic institutions with research expertise can produce results in months, not years.
For operations teams, the takeaway is concrete. Start with a specific pain point - regulatory compliance, inventory waste, or delivery accuracy - and measure the baseline before applying AI. The Carnegie Mellon projects each targeted a defined problem with clear metrics, which is why the results are worth studying.
Your membership also unlocks:
