Durance AI Joins the Siemens Cre8Ventures Digital Twin Marketplace

Reading Time: 2 minutes

We are pleased to announce that Durance AI has joined the Siemens Cre8Ventures Digital Twin Marketplace.

Durance AI is developing a next-generation neuromorphic vision processor designed specifically for autonomous mobile systems — including dual-use drones and Autonomous Mobile Robots (AMR).

The Problem: AI Is Power-Hungry

Deploying advanced AI in autonomous mobile systems remains a structural challenge.

Battery-powered platforms face a harsh trade-off: more AI capability often means dramatically reduced mission time

Running conventional AI on small drones can reduce flight time by around 50%. In GNSS-denied or contested environments, where onboard intelligence must replace external positioning systems, this becomes mission-critical.

Power efficiency is not a “nice to have”. It determines whether autonomy is viable.

What Durance AI Does Durance AI designs ultra-efficient edge AI hardware using neuromorphic computing architecture.

Their proprietary SPU (Spiking Processing Unit) enables:

~100x improvement in energy per inference
Real-time multi-target object detection
~500 mW power consumption
~20 TOPS compute performance

Unlike conventional accelerators, their architecture is optimised for event-based vision sensors, delivering:

Very low latency
Very low power consumption
High dynamic range
No motion blur

This places Durance AI in a rare performance envelope:

Compute >10 TOPS at <1W power.

In short: advanced vision intelligence without destroying battery life.

Why This Matters for GNSS-Denied Navigation & Perception

Within our first-of-a-kind sovereign GNSS-deprived Navigation & Perception subsystem, perception is foundational.

The primary risks in autonomous systems today are not mechanical. They are integration and subsystem risks:

Navigation Drift – drift under GNSS denial
Latency Coupling – delays within sensor fusion pipelines
Power & Thermal Constraints – SWaP limitations in autonomous platforms
Security Boundaries – protecting interfaces across subsystems

Durance AI strengthens the perception layer of this architecture.

Their technology directly supports:

Low-latency visual-inertial odometry
SLAM-based autonomous navigation
Multi-object detection in high-resolution imagery
Vision-based swarm intelligence

For dual-use drones and AMRs operating in contested or GPS-deprived environments, this capability enhances mission resilience while preserving endurance.

Sovereign Impact and EU Chips Act Alignment

The EU Chips Act emphasises:

Strengthening European semiconductor capability
Supporting deep-tech start-ups
Building resilient sovereign supply chains

Durance AI contributes to these objectives through three areas:

1. European Neuromorphic Semiconductor IP

Their SPLEAT neuromorphic IP is based on exclusive academic licensing and demonstrated through FPGA platforms.

2. Energy-Efficient Edge Compute for Strategic Domains

Drones, robotics, automotive and defence are sectors where energy-efficient AI compute is strategically critical.

3. Subsystem-Level Sovereign Integration

By integrating Durance AI into a transparent subsystem-level validation environment, reliance on opaque non-European AI hardware can be reduced.

This strengthens European capability at both the silicon level and system level.

Strengthening the Dual-Use Drone & AMR Campaign

The integration challenge in autonomy is increasingly about:

Compute efficiency
Sensor fusion latency
Thermal and power trade-offs
Mission-level validation before fabrication

Durance AI enhances the ecosystem by delivering:

Advanced neuromorphic perception
Swarm-ready visual intelligence
Ultra-low-power AI compute
Integration-ready subsystem capability

This directly enriches the sovereign GNSS-deprived navigation and perception initiative.

Looking Ahead

Durance AI’s participation in upcoming sovereign GNSS-denied workshops will provide defence organisations, industrial automation leaders and autonomy innovators with: