802.11bn
Ultra High Reliability. Draft 2.0 entered a new working-group ballot in August 2026. Reliability, latency under stress, mobility and consistent performance are the core story.
HAAM / Future networks / Wi-Fi 8
When Wi-Fi stops being a pipe.
Wi-Fi 8 is being designed around ultra-high reliability rather than a headline speed race. At the same time, IEEE is opening separate work on AI offload and intelligent networking. The interaction opportunity is to design what happens when local connectivity starts behaving like dependable computing infrastructure.
persondeviceaccess pointlocal AIspacecloud
Reliability changes behavior.
Wi-Fi 7 pushed peak throughput. Wi-Fi 8, IEEE 802.11bn, is formally an Ultra High Reliability project. The design target is a WLAN that behaves better in difficult real-world conditions, including crowded networks, coverage edges, mobility and handoffs.
That sounds like infrastructure. It becomes interaction design the moment a person expects a live AI conversation, spatial interface, wearable or remote control session to continue while they move through a building.
The experience should continue even when the network underneath it changes.
Three layers. Do not mix them up.
The interesting future is real, but the standards are at different stages. Keeping those stages visible makes the design research stronger rather than less ambitious.
AI Offload
IEEE created the AI Offload Study Group in March 2026 to investigate moving intensive AI inference toward Wi-Fi access points and other Wi-Fi edge-compute devices. The group approved a proposed P802.11bu project package in July.
WIN
The WLAN Intelligent Networking Study Group was approved in July 2026 to develop the project basis that IEEE says should form the foundation for Wi-Fi 9.
Design the room, not the router.
The useful mental model is not a faster box with antennas. It is a local environment where connectivity, nearby devices and potentially local compute cooperate around the person.
That creates a new interface question: what intelligence does this place offer, and under what rules can my devices use it?
Prototype 01
This Room Has Compute.
Instead of asking only for a Wi-Fi password, your device discovers what local capabilities a space offers, where processing happens and what happens to your data when you leave.
HAAM Studio / local networkCompute available in this room
Live translationLOCAL
Vision inferenceLOCAL
Room contextLOCAL
Large model fallbackCLOUD
Camera input is not retained. Local session context disappears when you leave this network.
Use local compute while I am here
Six interaction experiments to build now.
None of these requires waiting for commercial Wi-Fi 8. They can be prototyped with current phones, laptops, access points, local models and a small edge computer, then revisited as the standards mature.
Compute continuity
A task should survive as the computer doing it changes. Design the handoff when inference moves between glasses, phone, laptop, access point and local edge compute without forcing the person to restart anything.
Space intelligence permissions
If a room offers local vision, translation, storage or inference, people need a way to understand what exists, what stays local, what is retained and what they are choosing to use.
Experience intent
Applications should be able to express what the human experience needs: continuity, low latency, privacy, battery savings or graceful degradation. The network can optimize around intent rather than raw bandwidth alone.
Private local AI
Design AI that belongs to a home, studio, school, hospital or office and can answer locally without automatically sending sensitive context to a distant cloud service.
Cooperative access points
When multiple access points coordinate around the same experience, the user should perceive one continuous environment rather than a sequence of routers, passwords and invisible handoffs.
Network legibility
Make invisible infrastructure understandable without turning people into network engineers. Show just enough about reliability, locality, privacy and compute placement for a person to make a meaningful choice.
Local networks are only half the story.
Wi-Fi 8 is the near-term local layer. 6G raises the same questions at network and city scale, adding sensing, distributed compute and mobile infrastructure.
Continue to HAAM / 6G →Primary sources
Follow the standards, not the hype.
01IEEE 802.11 TGbn, Ultra High Reliability / Wi-Fi 8 project updates↗02IEEE P802.11bn project authorization request↗03IEEE 802.11 AI Offload Study Group↗04IEEE 802.11 Working Group, July 2026 WIN Study Group update↗05IEEE 802.11 AI/ML Standing Committee↗06Qualcomm, Wi-Fi 8 technologies powering ultra-high reliability↗
Prototype the interaction before the hardware is ordinary.
HAAM is exploring the human layer of reliable local networks: compute continuity, local AI, network intent, legibility and permissions for intelligent spaces.
Explore this with HAAM →