Zebra Technologies outlines a practical path for frontline AI adoption

Zebra Technologies says labor shortages and supply chain volatility are pushing enterprises to apply AI closer to frontline work.

Pressure on frontline operations is no longer limited to one industry or region. Labour shortages, rising customer demands, and supply chains that remain hard to predict are forcing many organizations to rethink how work gets done on the ground.

These themes were central at the Zebra Technologies Kick-Off APAC 2026 in Incheon, Korea, where the company shared how it is applying AI, sensing, and automation across frontline environments.

Tom Bianculli, Zebra Technologies’ chief technology officer, framed the discussion around a simple point: productivity depends on what happens where work actually takes place. That includes store aisles, factory floors, warehouses, loading docks, and delivery routes. Zebra’s tools have long been designed for these settings, beginning with barcode scanning and on-demand thermal printing, which still support many tracking systems today. Over time, that base has expanded to include rugged mobile devices, RFID, robotics, and machine vision.

RFID has become a growing focus as companies look for more consistent ways to track goods through supply chains. Zebra has worked with RFID for roughly two decades, but adoption has increased in recent years.

Around 20 billion RFID tags were shipped globally in 2020. That figure has climbed to an estimated 80 billion this year and is expected to pass 110 billion by 2028. As tagging becomes more common, companies are moving away from one-off scans toward continuous visibility that supports faster decisions.

Retail remains Zebra’s largest market, accounting for about 30% of revenue, followed by manufacturing and transportation and logistics. Manufacturing has seen strong growth over the past year, with Asia-Pacific playing a key role. Countries such as South Korea and India have been major contributors to that momentum.

Across these sectors, Bianculli said customers tend to face the same pressures. Labour is harder to find and more costly. Customer expectations continue to rise, especially around product choice and delivery speed. Supply chains remain unpredictable, shaped by economic and geopolitical factors. At the same time, organizations are under pressure to improve productivity to protect margins. Together, these forces are pushing companies beyond basic visibility toward what Zebra describes as intelligent operations.

“From an intelligent operations perspective, what it really means is bringing AI, data, and human expertise together to optimize workflows,” Bianculli said. “It’s about taking what people know about how work should be done and applying that consistently through automation.”

In manufacturing, this approach is already visible in quality inspection. Machine vision cameras along production lines capture images as products move past at speed. AI models review those images to flag defects or irregularities, using quality standards defined by human experts in advance. This reduces reliance on manual inspection and helps prevent faulty products from moving further down the supply chain, where issues become more costly to fix.

Warehouses and logistics operations follow a similar pattern. Mobile devices now act as sensors rather than simple scanning tools. They collect spatial and 3D data that can be combined with RFID to confirm pallet contents, track asset movement, and check that goods are where they should be. In delivery routes, AI-supported steps can help drivers locate the correct package more quickly or capture proof of delivery with less manual effort, saving time and reducing the risk of errors.

Retail has become one of the earliest areas to test these ideas. High staff turnover creates pressure to help workers learn tasks quickly and carry them out consistently. Many stores already rely on mobile devices for task assignment and communication, making it easier to introduce AI-based support that guides routine work or surfaces product and policy information when needed.

“What we’re seeing now is AI moving much closer to the frontline,” Bianculli said. “That’s where real-world data is created, and that’s where AI can have a very practical impact on day-to-day work.”

Many of these examples fall under what Bianculli described as physical AI. This includes fixed machine vision systems, RFID readers, and high-speed 3D imaging used in tasks such as picking, placing, and packaging. Together, mobile and fixed systems help create digital representations of physical spaces that support coordination and planning.

Bianculli cautioned that progress depends on focus. Data preparation remains a common challenge in AI projects, which is why he advises starting with specific workflows rather than trying to prepare all enterprise data at once. “Trying to harmonize all enterprise data at once rarely works,” he said. “When you focus on a specific workflow, you can identify exactly what data you need and start delivering value much faster.”

Early returns often come from time saved at the task level. Over time, cost avoidance becomes more important, such as catching shipment errors earlier or reducing penalties and rework. For many enterprise customers, a return on investment within 12 to 15 months is a typical expectation.

As these systems expand across networks of stores, factories, and distribution centres, the focus shifts from isolated improvements to coordinated execution. Applying insight consistently across large networks helps companies move toward more predictable operations.

Ryan Goh, who now leads Zebra’s Asia-Pacific business as well as its global OEM unit, said these pressures are becoming more visible across the region. “Businesses, especially in markets such as India and Japan, are under growing pressure from disrupted supply chains and ongoing labor shortages,” he said. “That’s driving stronger demand for intelligent automation at the frontline.”

As Bianculli framed it, the next phase of frontline operations is less about sweeping change and more about steady improvement—using visibility, AI, and human input together to make everyday work more reliable, one workflow at a time.