autonomous last-mile platform that enables delivery to a smart mailbox by drone, ground robot or human. Billions of dollars are lost to package theft, incorrectly addressed packages, and damage, notes Dan O’Toole, founder and CEO of Arrive AI. Smart mailboxes have the potential to limit those losses, providing a new element to the challenges of the last mile. Arrive’s smart mailboxes also include climate control and chain-of-custody measures, and the company’s system integrates with smart home devices such as lighting and security systems. A Growing Need For now, automated delivery is available only in pockets across the United States, but as automated delivery becomes more prevalent it will increase the need for smart mailboxes. Arrive was the rst to patent a mailbox capable of drone delivery—which the company calls Arrive Points, O’Toole notes. “One of our partners in India delivers about 200,000 packages a day via drone. When we have our Arrive Points in place there—we hope to have dozens there this calendar year—we’ll have a great use-case for our effectiveness,” O’Toole says. The addressable market for smart mailboxes is “huge,” O’Toole says. “In the United States alone, there are more than 160 million traditional mailboxes that haven’t been updated since 1858. We expect to continue re ning our product as we roll it out. “At scale, the data collection and additional features—emergency lighting, interaction with household IoT, video and advertising—are immense,” he says. “We see automated delivery as a necessary utility in the future for all consumers, retailers, and businesses.” 5 AI-based demand forecasting The use of AI in demand forecasting is helping retailers and their partners analyze vast amounts of data to make quicker and better-informed decisions. AI-based demand forecasting
o9 Solutions’ platform, which combines advanced analytics, scenario planning, and intuitive tools, uses AI to drive business planning across the enterprise—from demand forecasting to supply alignment.
transforms retail supply chains from reactive to responsive. “Traditional models rely on historical patterns—as long as the road ahead looks like the road behind you, you may be ne,” says Prashant Bhargava, senior vice president, product management, for the retail-CPG division of SymphonyAI. “These approaches break down when promotions shift, weather hits, or customer demand spikes. “AI doesn’t just predict, it adapts,” he adds. “It reads signals—from store inventory to external events—as they emerge and adjusts forecasts accordingly. AI-based demand forecasting delivers dramatically higher forecast accuracy, fewer stockouts and leaner, more ef cient inventory.” AI-based demand forecasting supports retailers who are “facing more volatility than ever,” notes Anjali Burkins, senior director of retail strategy, North America at o9 Solutions, a Dallas-based supply chain software company. “Demand patterns are shifting quickly, in uenced by factors such as local events, promotions and social media,” Burkins says. “At the same time, macro- level disruptions such as tariff changes and geopolitical tensions are creating new constraints and forcing teams to respond faster and more precisely. AI-based forecasting helps create a stronger foundation for planning in this environment.”
One common challenge teams face is struggling to pinpoint why they missed or exceeded plans. “Was it a demand surge in a certain region? A late delivery from a supplier?” Burkins says. “AI helps automate that root cause analysis and turns historical data into insights that are relevant for the next decision cycle. “Accurate forecasts allow teams to adjust plans earlier and align with supply chain partners,” she adds. “This could mean shifting order volumes, reallocating inventory, or triggering a scenario simulation to evaluate the best course of action. It enables planners to be proactive and collaborative, rather than reactive and siloed.” A New Way of Working Warehouse-level AI demand forecasting is in widespread use, while store-level adoption is catching up as retailers “realize that you can’t manage local nuances—customers, weather, traf c, seasonality—with historical averages,” Bhargava says. “The key advantage, however, is the ability to adjust plans and execute quickly across merchandising, supply chain, and stores as conditions change.” As Burkins notes, “The complexity of today’s environment demands a different way of working—one that brings together data, technology and people to support faster, more coordinated decisions.” n
36 Inbound Logistics • August 2025
Powered by FlippingBook