Edge Solution Summit 2026 Taipei

With equity in mind, digital literacy and AI education remain out of reach for many remote communities lacking reliable internet access. The challenge is how to enable meaningful digital and AI learning without dependence on cloud connectivity or data centers.
Intel's Edge Group addresses this with a purpose‑built, low‑power Edge AI solution based on Intel^® Core™ Ultra Series 3 (Panther Lake). The platform integrates compute, AI inference, graphic acceleration and multi‑user access in a single system, supporting up to four students with AI workloads running fully at the Edge.
This enables not only connected urban classrooms, but also off‑grid schools—transforming them into self‑sustained, AI‑enabled learning environments and extending digital literacy to students previously left behind.

Artificial Intelligence is transforming education by reshaping how learning is delivered, managed, and experienced. This presentation explores how AI powered intelligence enables personalized learning, data driven decision making, and inclusive educational environments. It examines intelligent analytics, adaptive learning systems, and smart education platforms addressing challenges such as learning disparities, limited resources, and digital skills gaps.
Through practical use cases, the session highlights how AI supports educators, enhances learning efficiency, and builds scalable educational ecosystems. Participants gain insights into AI's role in driving sustainable educational transformation, improving learning outcomes, expanding access, strengthening institutional effectiveness, and preparing learners and educators for future ready digital societies worldwide.

Ninety-five percent of enterprise AI investments fail to deliver meaningful returns. The 5% that succeed follow a repeatable playbook — and together, Iternal and Intel are putting it directly into the hands of enterprise leaders through the AI PC.
In this session, John Byron Hanby IV, best-selling author of The AI Strategy Blueprint — heralded by Fortune 500 CIOs and CAIOs as the definitive guide to AI transformation — reveals the frameworks separating AI leaders from laggards: the 10-20-70 Rule, Crawl-Walk-Run deployment, and ROI quantification validated across Fortune 500, defense, and classified environments. Discover how Iternal's AirgapAI on Intel AI PCs delivers 78x accuracy gains, turning strategy into measurable business outcomes your organization can execute immediately.

Artificial Intelligence offers powerful opportunities to enhance education, yet adoption is often limited by connectivity, security, and digital equity challenges.
This session introduces the C3 Micro‑Cloud, an edge‑based AI learning platform built in collaboration with Intel technologies to deliver digital content and generative AI capabilities in classroom environments with full, limited, or no internet access. Powered by Intel^® AI‑capable processors, the C3 Micro‑Cloud integrates local content, small language models, built‑in Wi‑Fi, and AI guardrails into a secure, plug‑and‑play solution. By enabling SAFE AI—Secure, Appropriate, Functionally specific, and Economically feasible—the platform supports personalized learning while helping schools standardize experiences and responsibly bridge the digital divide.

This session focuses on key trends shaping retail edge devices, including AI adoption, OS trends,  and rising manageability and security needs. We will show how the Intel Edge roadmap maps to these trends and where it enables scalable retail deployments. The update will also highlight capabilities of the Retail AI Suite and conclude with what’s next for ISVs, including priority areas for alignment and collaboration across the retail ecosystem.

Deploying AI at the retail edge isn't a model problem — it's a lifecycle and orchestration problem. From POS to kiosk to digital signage, every device runs a different OS, ships on a different update cadence, and carries a different security posture. Without a unified way to provision, update, monitor, and recover that fleet, AI models can't be trusted in production. And untrusted AI at scale isn't just an ops headache — it's a compliance and liability risk.
Esper's session introduces a four-step orchestration playbook built for heterogeneous, multi-OS retail environments:
(1) unifying fragmented device fleets across operating systems under a single control plane;
(2) surfacing real-time edge telemetry that feeds directly into operational AI pipelines;
(3) integrating secure device lifecycle data with customer and inventory systems for full-context, trustworthy AI decisions;
(4) and automating model deployment, rollback, and recovery with DevOps-grade reliability and auditability.
Built on Intel-validated silicon, Esper's platform provides the deterministic compute, secure boot, and end-to-end lifecycle control that production AI actually demands — transforming legacy retail infrastructure into AI-ready assets that operators can deploy with confidence, update without risk, and audit without friction.

Digital signage is no longer just about displaying content. It is becoming the intelligence layer of modern retail. In this session, we explore how Edge AI is transforming traditional signage into a connected retail ecosystem.
Beyond promotions, retailers can integrate AI-powered self-ordering, smart wayfinding, drive-thru automation, and data-driven content personalization, all running efficiently at the edge. Powered by Intel^® Edge AI technologies, these solutions enable faster decision-making, improved operational efficiency, and stronger customer engagement.
“Beyond Digital Signage” represents the shift from screens to smart retail infrastructure, where physical spaces become intelligent, responsive, and revenue-generating environments.

Public transportation systems face a dual crisis: rising passenger expectations and critical staffing shortages. This session unveils how Pantheon Lab leverages the Intel^® Core™ Ultra Series 3 (Panther Lake) to deploy autonomous 'Agentic Digital Concierges' at the edge. Moving beyond static kiosks, these agentic humans utilize local NPU 5.0-driven emotion detection and OpenVINO™-optimized LLMs to resolve complex wayfinding, ticketing, and emergency scenarios in real-time.
We present validated case studies from SBS Transit (Singapore) and MTR (Hong Kong), which demonstrated a 50% reduction in staffing costs and a 30% drop in passenger complaints. Attendees will learn how integrating the Metro AI Suite's agentic frameworks with Pantheon's 'Digital Human as a Service' creates a scalable, privacy-first infrastructure that modernizes the passenger journey while delivering double-digit ROI.

Optimize Crowd Management with WaitTime AI explores how organizations can transform camera infrastructure into real-time operational intelligence that drives efficiency, safety, and revenue. This session will demonstrate how WaitTime's patented, anonymous AI analyzes live video streams (RTSP) to measure crowd flow, queue lengths, and occupancy levels—at incredibly high volumes.
Attendees will learn how retail environments, and large public spaces are using WaitTime to reduce wait times, improve throughput, and enhance the customer experience while simultaneously unlocking new revenue opportunities. The presentation will also highlight how integrating WaitTime into existing security and infrastructure ecosystems enables a shift from passive monitoring to proactive decision-making.
Ultimately, this session will show how organizations can move beyond simply observing crowds to truly understanding and optimizing them—turning everyday operations into measurable, data-driven outcomes.

This presentation explores the collaboration between Sensory AI and Intel in pioneering the "New Retail" frontier. We feature Ella, the world's first fully autonomous robotic barista, as a primary case study in digital transformation.
By integrating Intel^® Core™ processor, Sensory AI utilizes advanced edge computing and computer vision to ensure safety, quality control, and operational efficiency in a 24/7 unmanned environment. We will present how this "Physical AI" architecture addresses critical industry pain points—such as labor shortages and rising rents—while delivering a hyper-personalized customer experience through mobile integration.

As cities evolve from connected infrastructure to intelligent systems, on-device AI is becoming the foundational layer enabling scalable, secure solutions across safety, security, and transportation. This session introduces Intel Metro AI Suite, an application framework that accelerates partner development of Vision AI and Gen AI solutions through reference implementations, benchmarking tools, and validated system designs.
Attendees will explore a real-world deployment example from Intel AI and Smart Cities ecosystem, which integrated ~18,000 cameras with License Plate Recognition, video analytics, and investigation tools—achieving an 11.83% reduction in incident rate and 99% detection accuracy.
Designed for OXMs, SIs, and ISVs, this session provides architectural guidance and ecosystem alignment strategies to help partners build and monetize smart city solutions on Intel-based platforms.

Smart cities promise seamless connectivity — but most implementations trade privacy for convenience, relying on centralized cloud infrastructure that struggles with real-time personalization. IrisGo introduces a new paradigm: the AI PC as an intelligent edge node that bridges personal context with urban services.
Built on Intel's AIPC platform, IrisGo's on-device AI engine continuously learns user preferences, behaviors, and daily routines —without sending sensitive data to the cloud. When integrated with smart home ecosystems, this personal context layer extends naturally into smart city infrastructure: energy management that adapts to your habits, city services that respond to your schedule, and ambient environments that know you before you arrive — all processed locally, on-device.
This session explores how IrisGo's Skills Marketplace enables developers to build privacy-first smart city applications on Intel AIPC, creating a decentralized intelligence network where every AI PC becomes a citizen node in the urban ecosystem.

HoloXit introduces a Unified Intelligence Platform that transforms multi-source urban data into real-time operational intelligence for Smart City environments. As cities deploy increasing numbers of cameras, drones, and IoT sensors, the challenge is no longer data collection but processing, integration, and decision-making at scale.
Built on Intel^® Edge AI technologies, the HoloXit Appliance enables high-performance AI inference directly at the edge using Intel processors and optimized AI frameworks. The platform integrates AI video analytics, drone-based surveillance, social media intelligence, and AI-assisted field reporting into a unified operational dashboard for city command centers. By combining edge computing, real-time analytics, and scalable architecture, HoloXit allows cities to process data closer to the source, reducing latency and improving situational awareness. The solution supports key Smart City applications including traffic monitoring, public safety, infrastructure protection, and incident response.
Through an open and scalable ecosystem approach, HoloXit demonstrates how Intel-powered edge AI platforms can accelerate the deployment of intelligent, responsive, and data-driven urban infrastructure.

Managing modern buildings requires more than traditional facility systems. With increasing complexity in infrastructure, energy consumption, and asset management, organizations are turning to AI and IoT technologies to create more connected and efficient environments.
This session explores how ALEF 360 supports the digital transformation of the built environment by bringing together IoT data, operational workflows, and analytics within a single platform. By connecting building systems, assets, and environmental monitoring, ALEF 360 enables facility teams to gain real-time insights into operational performance.
Intel^® Edge AI plays a key role in enabling this transformation. Edge computing allows AI models to process data directly from devices and sensors, improving response time and reducing latency. This enables automated alerts, predictive insights, and more responsive building operations.
By combining ALEF 360 with Intel's edge computing capabilities, organizations can transform how buildings are operated and maintained—improving efficiency, supporting sustainability goals, and enabling a smarter built environment.

OKEAN is an AI-powered Video Intelligence Platform purpose-built for smart city and enterprise environments, delivering real-time insights from video streams at the edge. Designed for scalability and operational simplicity, OKEAN transforms standard camera infrastructure into intelligent sensing nodes — without replacing existing deployments.
At its core, OKEAN offers purpose-built applications including TrafficMon for traffic monitoring and analytics, and DeepVue for deep vision-based detection and behavioral insights. These applications run seamlessly on Intel Core Ultra-powered edge devices, leveraging the integrated NPU for efficient AI inference through Intel's OpenVINO and OpenEdge software stack.
OKEAN's Central Management System (CMS) runs on Intel Xeon-powered servers, enabling centralized orchestration and analytics aggregation, at scale — creating a cohesive full-stack deployment across both edge and server tiers.
OKEAN's architecture prioritizes low latency, privacy-first processing, and minimal cloud dependency — making it ideal for deployments where real-time decisions matter. From traffic management to security and safety analytics, OKEAN delivers actionable intelligence precisely where it is needed — at the edge.

This presentation highlights a Honeypot & Vulnerability Scan Managed Service that helps organizations detect cyber threats early and reduce security risk. Honeypots act as decoy systems to identify real attack activity without affecting production environments, while automated vulnerability scans continuously uncover system weaknesses that require remediation. Delivered through 24/7 managed monitoring, the service provides timely alerts, clear risk visibility, and actionable recommendations. By combining live threat detection with continuous risk assessment, organizations can improve security posture, shorten response time, and minimize the likelihood and impact of cyber incidents—without increasing operational complexity.

Manufacturing is an input-transformation-output complex environment. From automation, vision inspection machines, robotics at all kinds, to operation governance processes involving production flow planning, capacity simulation, and processes governance; all these dominate solution space we deal in manufacturing environment. Added on top nowadays is the AI revolution in factory environment. The factory floor is no longer just about hardware - it's about intelligent transformation. Intel processors never absent from this domain industry. We continue on as a strong industry brand and critical compute foundation to commit factory managers a reliable and trustable computer platform.
Come to join us at the event to learn Intel portfolios dedicated to manufacturing environment, and listen to our global software solution partners on their turnkey ready-to-deploy solution innovate around Intel portfolios to address use cases in factories.

Semiconductor manufacturing is entering an era in which traditional automation is no longer sufficient.
The presentation will examine what it takes to mass deploy AI agents at the edge—systems that can perceive equipment states, make contextual decisions, and execute actions safely in real time. It will describe how organizations can scale from isolated proof of concept deployments to fleets of agents operating across heterogeneous tools, legacy interfaces, and globally distributed sites. A practical reference architecture for edge native agents will be presented, including perception layers that interpret tool outputs, policy governed decisioning, closed loop control, and integration with MES/FDX/SPC systems and SECS/GEM environments.
The discussion will address the operational constraints most critical to high volume manufacturing: deterministic latency, availability, cybersecurity, change control, validation, and auditability.
Finally, a rollout roadmap will be proposed—standardizing reusable agent “skills,” establishing lifecycle management, and implementing governance—so edge AI can become a dependable industrial layer for yield improvement, OEE gains, and operational resilience.

AI vision inspection is no longer optional in manufacturing. It is essential for quality, productivity, and competitiveness. Yet deploying AI reliably in real industrial environments remains a challenge. While GPUs are often treated as the default solution for AI workloads, they are not always the most practical choice for edge deployment due to cost, power consumption, thermal constraints, and system complexity.
At Neurocle, we provide an Auto Deep Learning–based platform that enables industrial experts to build and deploy their own vision models without requiring deep AI expertise. From this perspective, hardware flexibility is critical: manufacturers should be able to select the platform that best fits their operational environment.
This session explores a more practical approach to industrial Edge AI by leveraging Intel's heterogeneous computing platforms—including CPU, integrated GPU, and NPU. By integrating Intel OpenVINO into our runtime, we optimize performance across diverse Intel-based systems, enabling scalable, efficient, and industry-ready AI deployment beyond proof-of-concept.

This presentation showcases the integration of MVTec's machine vision software with Intel platforms. MVTec demonstrates how machine vision is applied in real industrial and manufacturing scenarios. It also explains the AI² accelerator interface of OpenVINO^TM and how it enhances the performance of MVTec software on Intel hardware. In addition, the presentation highlights MVTec's deep learning solution, with benchmark results on Intel's latest platform illustrating improved efficiency and accuracy. These results demonstrate how Intel hardware enables faster processing and more reliable machine vision outcomes.

COSAP is an AI-powered command and operations platform purpose-built to transform manufacturing businesses through real-time intelligence and data-driven decision-making.
Designed for the smart factory era, COSAP connects ERP systems, production data, inventory, and supply chain operations into a single unified intelligence hub, eliminating the disconnected silos that cost manufacturers time and money.
Its Auto Ontology engine automatically maps business relationships and translates raw data into actionable insights, while its multi-layer AI risk detection monitors for fraud, stock shortages, supplier failures, and compliance issues before they escalate.
Through conversational AI, plant managers and operations teams can query live production data in natural language, no SQL required, and receive answers in seconds.
With purpose-built modules for BOM management, production analytics, inventory control, and quality checks, COSAP empowers manufacturers to respond faster, operate leaner, and build factories that are truly AI-ready.

This presentation explores how Smart Manufacturing—using AI, IIoT, digital twins, and edge computing—drives competitive advantage amid supply chain volatility, talent shortages, and geopolitical shifts. It covers real-time analytics to reduce downtime, autonomous quality inspection to lower labor dependency, and adaptive systems for nearshoring.
Attendees gain practical pathways to turn disruptions into opportunities, aligning Industry 4.0 with strategic goals for resilience, efficiency, and sustainable growth.

Robotics is evolving from isolated automation to intelligent, adaptive systems operating in real-world environments. The challenge is no longer building AI models. It is deploying, scaling, and managing these systems reliably across diverse use cases.
In this session, we will outline Intel’s robotics strategy and how we are enabling partners to move from pilot to production using the Open Edge Platform (OEP). By combining Intel hardware, software, and an open ecosystem approach, OEP provides a consistent foundation to develop, deploy, and scale robotics AI across environments.
By combining Intel hardware, software, and an open ecosystem of partners, we help accelerate time to deployment and reduce complexity across the stack. Attendees will gain insight into how Intel is enabling partners to bring robotics AI to production faster, and how OEMs, ISVs, and system integrators can collaborate to build production-ready solutions and unlock new opportunities.

As advanced AI models bring perception and reasoning to Physical AI, a critical challenge remains: executing AI decisions with deterministic, low-latency, and synchronized motion in real-world systems.
WMX is a software-defined real-time motion control stack optimized for Intel^® Core™ Ultra 3 Series processors, delivering microsecond-level synchronization and scalable multi-axis control directly on IPC platforms—without dedicated motion control hardware.
By leveraging Intel^® Time Coordinated Computing (TCC), WMX maximizes deterministic scheduling and predictable real-time performance on Intel architecture. This enables AI inference, planning, and precision motion control to operate cohesively on a unified Intel edge platform—reducing hardware complexity while improving reliability and accelerating deployment.
This presentation will demonstrate how Intel^® Core™ Ultra 3 and WMX, integrated with Intel^® Open Edge Platform, enable scalable, industrial-grade Physical AI for next-generation robotics systems on a single edge platform.

Circulus introduces the "Universal AI Brain," a groundbreaking on-device Physical AI solution optimized for the Intel Panther Lake platform.
Our architecture maximizes Intel's tri-compute synergy：
1.NPU for low-power, real-time vision and privacy-centric de-identification.
2.GPU for complex Vision Language Model processing.
3.CPU for seamless system orchestration.
This hardware-agnostic intelligence layer enables unified control across diverse platforms, including South Korea's national humanoid MCR-1 and the Unitree G1. In 1H 2026, we are delivering these Intel-optimized solutions to government-led Physical AI project. By bridging high-performance silicon with scalable robotic logic.

Oversonic Robotics is redefining the role of humanoid solutions within the rigorous demands of the healthcare and industrial sectors.
This session explores how the integration of Intel^® Core™ Ultra Series 3 processors serves as the catalyst for true autonomous cognition at the edge. By leveraging the OpenVINO™ toolkit, Oversonic intelligently distributes sophisticated AI models across heterogeneous CPU, GPU, and NPU engines to maximize operational efficiency.This localized execution directly solves industry-specific challenges. In healthcare, the architecture ensures strict data privacy by keeping sensitive information on-device, while significantly extending battery life for long-shift reliability. For industrial environments, the result is ultra-low latency and peak performance in complex, dynamic workspaces.
Attendees will discover how the fusion of Oversonic's humanoid design and Intel's cutting-edge silicon delivers robots that are more responsive, secure, and energy-efficient than ever before.

Physical AI has captured enormous attention, but turning robotic intelligence into something useful in the real world still breaks down in the same place: not model ambition, but operational reality.
Teams can often imagine the application, train the model, and even produce a compelling demo, yet still struggle to build systems that are responsive, repeatable, practical to deploy, and fast to iterate on.
This session explores what actually has to come together to make physical AI useful beyond the lab, including embodiment choices, perception and control at the edge, latency, system integration, workflow design, and the speed at which teams can get from concept to working system.
The goal is to look past robotics theater and focus on the real bottlenecks that determine whether an idea becomes a usable capability.
This presentation is for organizations trying to move physical AI from impressive concept to practical development, evaluation, and deployment.