Inside Anduril and Meta’s quest to make smart glasses for warfare

· Source: MIT Technology Review · Field: Technology & Digital — Artificial Intelligence & Machine Learning, Robotics & Autonomous Systems, Emerging Technologies & Innovation · Depth: Intermediate, medium

Summary

Defense-tech company Anduril, in collaboration with Meta, is prototyping augmented-reality headsets for military use, aiming to optimize soldiers as "weapons systems" through integrated vision and decision-making. Two projects are underway: the Army's Soldier Born Mission Command (SBMC), a $159 million contract with Meta for AR glasses attached to existing helmets, and Anduril's self-funded EagleEye, an integrated helmet and headset combo. These systems, still years from production, envision soldiers ordering drone strikes via eye-tracking and voice commands, overlaying critical information like maps, drone locations, and AI-driven target recognition. Anduril's Lattice software, which the Army is integrating across its infrastructure with a $20 billion contract, will serve as the engine, processing data from various military hardware. The prototypes are also testing new digital night vision using generative AI and machine learning, and must overcome challenges like environmental durability, weight, battery life, and local AI processing without 5G.

Key takeaway

For defense contractors and military strategists evaluating next-generation soldier systems, Anduril's AR headset prototypes with Meta highlight a significant shift towards AI-augmented human-machine teaming. You should prioritize robust, intuitive interfaces that minimize cognitive load and ensure local AI processing for disconnected environments, as soldier adoption hinges on seamless integration and proven utility in high-stress combat scenarios, not just technical feasibility.

Key insights

Anduril and Meta are developing AR headsets to integrate soldiers and drones into a unified, AI-enhanced weapons system.

Principles

Method

The system uses large language models (Gemini, Llama, Claude) to translate soldier speech into commands, integrates data via Anduril's Lattice software, and employs eye-tracking for multi-step task execution.

In practice

Topics

Best for: CTO, Executive, Investor, AI Engineer, Computer Vision Engineer, AI Product Manager

Related on AIssential

Open in AIssential →

Editorial summary, takeaway, and curation by AIssential. Original article published by MIT Technology Review.