Wittra·Nida Tech Sweden·Physical AI
Investor briefing · 2026
The Missing Layer
of Physical AI.
The infrastructure every robot will need. Twelve years of RF engineering. 80+ patents. Proven in Europe's hardest industries — ready for the world's Physical AI platforms.
01 / 19
The moment·The Problem·Wittra layer·Solution·The moat·Go-to-market·Potential·Plan
The moment is now
Autonomous robots are
leaving the lab — and entering
the real world at scale.
2024–2026: Figure, Boston Dynamics, Tesla Optimus, Agility and Unitree begin commercial rollout of humanoid robots.
Physical AI is the fastest-scaling category in tech — factories, warehouses, airports and cities are next.
The window to own the layer they all depend on is opening — right now.
Wittra · The moment is now02 / 19
The moment·The Problem·Wittra layer·Solution·The moat·Go-to-market·Potential·Plan
The ChatGPT moment of Physical AI
Intelligence is
leaving the screen — and it can't
navigate the real world alone.
Foundation models can reason, plan and see. They cannot share reality.
Every robot, drone and AGV still operates from its own isolated sensors — in its own coordinate system.
The bottleneck is no longer intelligence. It's infrastructure.
Wittra · The Missing Layer of Physical AI03 / 19
The moment·The Problem·Wittra layer·Solution·The moat·Go-to-market·Potential·Plan
Disconnected chaos vs. one common layer
Many networks. Many systems. No unity.
Today
Disconnected chaos — 8+ parallel networks per site.
- Wired (Ethernet, PLC, SCADA, DCS)
- Industrial Wi-Fi · 5G / 4G · LoRaWAN
- Zigbee · Bluetooth · Mioty · Camera systems
- Double installations, higher cost, long lead times
- Data locked in silos, limited scalability
- Hard to reach full visibility or real insight
With Wittra
One network. One platform. Full control.
- One common structure for all data flows
- TCP / Ethernet, 5G, LoRaWAN, UWB, Zigbee, Mioty
- All communication on one unified fabric
- Lower installation cost, more reliable operation
- Full real-time visibility across the site
- Ready for edge AI and robotic coordination
Wittra · The Missing Layer of Physical AI04 / 19
The moment·The Problem·Wittra layer·Solution·The moat·Go-to-market·Potential·Plan
The Wittra layer · where it's already running
Steel. Mining. Airports.
Real sites — already live
on Wittra.
These are not renders. Every image is a real environment where Wittra is deployed today — molten metal, deep tunnels, dense terminals, high-noise industrial floors.
The exact places where GPS and Wi-Fi give up — and where Physical AI is being deployed first. Twelve years of R&D against real-world physics.
Wittra · Live in the real world05 / 19
The moment·The Problem·Wittra layer·Solution·The moat·Go-to-market·Potential·Plan
The Wittra layer · live at Sweden's largest airport
Arlanda runs
on the Wittra layer.
Aircraft, ground vehicles, autonomous wheelchairs, staff, tools and assets — all seen inside one coordinate system.
Wittra is the shared-reality layer Swedavia uses to coordinate machines and people across the terminal.
Wittra · Arlanda / Swedavia · Live deployment06 / 19
The moment·The Problem·Wittra layer·Solution·The moat·Go-to-market·Potential·Plan
One live view. One coordinate system.
Every asset,
on one map.
Sub-metre indoor positioning · continuous telemetry · one platform for tracking, maintenance and orchestration.
184 devices tracked live · airport-scale · no GPS · no Wi-Fi dependency.
Wittra · Arlanda dashboard · live07 / 19
The moment·The Problem·Wittra layer·Solution·The moat·Go-to-market·Potential·Plan
The first step · already in service
Arlanda already runs
autonomous wheelchairs
on the Wittra network.
A fleet of autonomous chairs coordinated through Wittra — sharing one map, one path, one intent.
This is the first step. Autonomous machines using the Wittra layer to move safely through a public, live environment — today.
Wittra · Step 1 · Autonomous wheelchairs at Arlanda08 / 19
The moment·The Problem·Wittra layer·Solution·The moat·Go-to-market·Potential·Plan
Back to the problem · the next step
Next: robots at work in these
same environments — but first
the navigation problem must be solved.
Those already on the Wittra network have taken the first step — one map, one coordinate system, sub-metre indoor positioning.
But robots need far more than a fix: shared reality, deterministic timing, orchestration across fleets, fallback across radios.
It's more complicated than that — and we've solved it. The next slides show how.
Wittra · From first step to full navigation09 / 19
The moment·The Problem·Wittra layer·Solution·The moat·Go-to-market·Potential·Plan
Live inference · Wittra × Cosmos × swarmCORTEX
Ten minds. One mission.
Watch them think.
Wittra fixes → Cosmos 3 reasoning → swarmCORTEX directives. The chain that turns raw physics into coordinated decision.
+5.9% ETA · +12.1% orchestration · +4.7% live vLLM — measured, not modeled.
Live inference stream · COSMOSSYSTOWN-SWARMHIVE · 100% election agreement
Wittra · Live Physical AI demonstration10 / 19
The moment·The Problem·Wittra layer·Solution·The moat·Go-to-market·Potential·Plan
Solution · seconds and metres only
22.4% faster — under
real physics. Proven,
not promised.
cameras only · alone · 173.38s → +systownCOSMOS3 · one by one · 164.23s → +swarmCORTEX3 · as one mind · 134.54s → theoretical optimum · 134.18s
22.4–23.8%
Lift under real kinetic physics
shift wave · 60 never-seen missions · ≤1.7% from ceiling
7 / 9
Certified missions
demo mission = exact MILP optimum
84,631
Live Cosmos 3 calls
0.14% errors · 100% election agreement
Wittra · The map alone: +5.2%. The mind on top: +18.1%.11 / 19
The moment·The Problem·Wittra layer·Solution·The moat·Go-to-market·Potential·Plan
The Horizon
One shared frame
for everything that moves.
HumanoidsHumansDronesVesselsVehicles
Live Wittra hardware — a credentials switch away.
Native embedding injection — the modality goes deeper.
Cosmos 4 lands — EVOLUTION 4 is ready.
In plain words: the next versions are already on the runway.
Wittra · The Horizon12 / 19
The moment·The Problem·Wittra layer·Solution·The moat·Go-to-market·Potential·Plan
Scale
From one warehouse
to the world's floor.
Ten robots today. Ten thousand across Manhattan next. Four hundred thousand by 2030. Each rung earned, never promised.
Manhattan — 10,000 agents, one frame · PLANNED
Federation — 500 gateways in sync · PLANNED
2030 — 400,000 agents, one city mind · PLANNED
In plain words: pass the test. Take the step.
Wittra · Scale roadmap13 / 19
The moment·The Problem·Wittra layer·Solution·The moat·Go-to-market·Potential·Plan
Where Wittra sits
Every giant is missing the same layer.
| Company | Strength | What they lack | How Wittra completes them |
|---|---|---|---|
| NVIDIA | AI compute, Omniverse, Cosmos, Isaac | No own industrial infrastructure or real-time network | Delivers real sensor data into AI and digital twins |
| Tesla | Robots, autonomy, AI | Only within its own ecosystem | Makes factories and infrastructure AI-ready, vendor-agnostic |
| Figure AI | Humanoid robots | Needs a connected physical environment | Gives robots real-time information from the entire facility |
| Boston Dynamics | Mobile robotics | Limited integration with the rest of OT | Connects robots to the rest of the factory |
| Siemens | Automation and Digital Twin | Many separate networks and systems | One common communication layer across every technology |
| ABB | Robotics and automation | Focus on their own systems | Neutral infrastructure between every vendor |
| Palantir | AI platform and analytics | Dependent on external data sources | Delivers structured real-time data from the edge |
Wittra · The neutral layer every giant needs14 / 19
The moment·The Problem·Wittra layer·Solution·The moat·Go-to-market·Potential·Plan
Go to market
We ride the giants' rails.
Ericsson
Siemens
NVIDIA
Ericsson — global 5G rollout channel & joint industrial deployments.
Siemens — automation and Digital Twin footprint across factories.
NVIDIA — Cosmos / Omniverse integration for Physical AI stacks.
Wittra · Partner-led go-to-market15 / 19
The moment·The Problem·Wittra layer·Solution·The moat·Go-to-market·Potential·Plan
Wittra base case · ARR by cohort · year-end (K EUR)
From €280K to €32M in five years.
Wittra · Base case ARR 2026–203116 / 19
Base case·Robotic upside·Group·Team
However — this is just the beginning
Wittra's potential in a $140bn market — we take 1–2%.
Wittra base case ARR (€0.3M → €32M)
Wittra potential — 1–2% of a $140bn market
A simple robot calculation
1
A robot costs €180 000
2
Depreciated over 3 years
= €60 000 per year, per robot
3
Make it 20% more efficient
= €12 000 saved per year, per robot
4
Wittra captures 2% of that gain
≈ €240 per robot per year
5
10 million robots by 2031
Conservative global forecast
Wittra · Our potential in a $140bn market — 1–2% capture17 / 19
Robotic upside·Group structure·Team
How the group is built
One IP core. Two go-to-markets.
Parent · IP holder
Nida Tech Sweden AB
80+ patents · patented RF network fabric · 12 yrs R&D · licenses the Missing Layer to both operating companies
↓ LICENSES IP TO ↓
Sister · Proof & cash flow
Wittra Industry
Deploys the layer in the world's hardest environments — steel, mining, airports, heavy manufacturing. Generates revenue and real-world telemetry.
Sister · Physical AI platform
Wittra AI Lab
Turns the layer into the shared-reality substrate for every autonomous platform — humanoids, AMRs, drones, vehicles. Software-first, global.
Wittra · Group structure18 / 19
Group·Team
The team
The minds behind the layer.
Management
CEO
Håkan Dackefjord
CTO
Warwick Taws
40 yrs · patents · B2B architecture
CEO Systown
Patrik Johansson
Interim CFO
Carl-Fredrik Morander
Commercial · Wittra Industry
CCO
Magnus Öhlund
CMO
Patrik Wigelius
Adj. Prof.
Johan Kullander
Project Mgmt
Eric Jacobsson
Tech Development · Wittra + Systown
Tech Lead
Ayberk Özgür
PhD EPFL · embedded & robotics
Tech Lead
Adrian Caragea
MSc · 6LoWPAN, CoAP, MQTT
Tech Lead
Beshr Al Nahas
PhD Chalmers · wireless sensor nets
Sw & Research
Pablo Corbalán
PhD · UWB & positioning · 14 patents
Product Owner
Spandan Dey
MSc KTH · firmware & IIoT
Systown CTO
Pontus Sahlberg
Systown
Kristian Pettersson
Edge AI System Lead
Systown
Ella Tojit
Physical AI Platform Lead
Systown
Edwin Andermyr
Agentic Platform Lead
Systown
Oscar Hansson
Agentic Physical AI Lead
+ Team
10 engineers
Deniz, Jimmy, Martin, Nagarajan, Rezaur, Tobias, Verdant…
Wittra · Team Wittra + Systown19 / 19