Our third year in Taipei, the hub of the world’s technology supply chain. EDGE AI Taipei 2026 will bring the knowledge of the worldwide EDGE AI ecosystem into Taipei to educate, connect, and accelerate the edge AI economy in Taiwan. Attendees will engage directly with leaders in edge AI from outside Taiwan as well as inside, and be exposed to opportunities to accelerate their business, partnerships, careers and technology knowledge at a global level.
This is an in-person event, which will be recorded. In addition to in person talks we have opened up an option for a pre-recorded talk for broadcast on the EDGE AI FOUNDATION YouTube channel.
As always , direct product pitches will not be accepted. Please focus your content on educating the community on a key technology topic for edge AI.
KEY TOPIC AREAS
- Resilient Edge AI
Innovative solutions that ensure edge AI performance and reliability in the most challenging and mission-critical environments. This track explores the intersection of hardware, software, and systems engineering to deliver resilient, energy-efficient intelligence from the data center to the extreme edge.
- Energy efficiency and resiliency from data center to edge
- Drones and autonomous systems on land, sea and air
- Defense Applications
- DDIL communication systems for edge AI
- Disaster preparedness and recovery
- Mission Critical SW Worklfows
2. Industry Transformation with Edge AI
Case studies or industry transformation using edge AI technologies in real world deployments. For example, the use of edge AI and cloud AI in industries such as:
- Manufacturing
- Agriculture
- Med Tech
- Retail
3. Solving the Commercialization Gap
Share successful edge AI implementations, lessons learned in commercialization, support and deployment, such as:
- Workload orchestration
- MLOps
- Commercialization challenges
- Supportability and Maintenance
- Cloud to edge architectures
4. Physical AI, Robotics & Autonomous Systems
Explore how edge AI enables intelligent machines through perception, reasoning, decision-making, and control in the physical world, including:
- Robotics and autonomous systems
- AI for drones and mobility
- Dexterity and manipulation
- Human-machine collaboration
- Safety and trust in autonomous systems
5. Edge AI Silicon & Systems
Advances in AI accelerators, NPUs, embedded platforms, edge infrastructure, and hardware-software co-design, including:
- Geopolitical issues regarding export controls and other policy challenges
- Model specific AI accelerators
- Memory and storage innovations
- Heterogeneous computing
- Hardware-software co-design with AI
6. Edge Generative AI & Agentic Systems
Foundation models, multimodal AI, intelligent agents, on-device AI, and emerging Generative edge AI applications, including:
- Small language models (SLMs)
- Multimodal AI
- Agentic AI systems
- On-device inference
- Personalization and privacy
- Synthetic data and model adaptation
7. Edge AI Development & Deployment
Tools, frameworks, model optimization,, benchmarking, and best practices for building and scaling edge AI solutions, including:
- Model optimization and compression
- AI compilers and runtimes
- Profiling and debugging
- Benchmarking and performance evaluation
- Development frameworks and toolchains
8. Security, Safety from Metal to Cloud
Computer vision, sensor fusion, multimodal sensing, AI security, privacy, reliability, and trustworthy Edge AI systems, including:
- AI model security
- Communication security supporting edge compute
- Adversarial attacks and defenses
- AI safety and reliability
- Explainability and transparency
- Secure deployment and lifecycle management
Submission Instructions & Review
SUBMIT YOUR ABSTRACT HERE
DEADLINE August 15,2026.
Please also provide a link to a 30 second video explaining your pitch – this will be used for evaluation and also to promote your talk once accepted!
Contributors will be notified by August 31.
Session Formats
Consider the following formats for your content:
-
- Workshop (60 minutes): Interactive hands-on format to develop solutions and provide learning experiences
- Technology Talk (20min ( 15 min content + 5min Q&A): Traditional presentation format on an engineering topic
- Panel discussion – strategy/industry/policy discussion
- Fireside Chat – one on one industry or strategy discussion
- Recording: 30 minutes recording of your presentation for our YouTube channel, to be released with event recording videos
- Poster Presentation: Visual display of research finding to be presented at the event