June 23, 2025

Auki community update recap: Jun 23, 2025

Post-SuperAI: Humanoids, Hybrid Robotics, and a Decentralized AI Future

We've just returned from nearly a month on the road, showcasing our innovations at major conferences and diving deep into the future of robotics and AI. Here are some highlights and insights from our journey.

Success at AWE and SuperAI

Our recent appearance at the Augmented World Expo (AWE) was a big success. We proudly won two awards:

  • Best Enterprise Solution for Cactus
  • Best Web3 Implementation for the Posemesh

Immediately following AWE, we headed to SuperAI in Singapore, where Nils moderated a key panel featuring leading humanoid robotics companies Unitree and Persona AI, discussing crucial questions about how humanoids will come to market.

Humanoid Robots: General AI vs. App Store Model

One intriguing debate that emerged was whether humanoid robots will run on a single general-purpose AI or use an "app store" model, with specialized applications handling specific tasks. We're firmly in the "app store" camp, believing it's the most realistic model in the near term. Due to significant hardware differences between robots, creating robot-agnostic manipulation apps remains challenging.

Our position was well-received, underscoring our perspective that robots will require distinct software layers—particularly separate layers for locomotion, manipulation, and application logic—to ensure cross-robot interoperability.

Auki’s Vision: Six Layers of the Humanoid Stack

Nils also gave a keynote presenting our structured vision for the humanoid stack, emphasizing six distinct layers:

  1. Locomotion
  2. Manipulation
  3. Perception
  4. Mapping
  5. Positioning
  6. Application

Each layer must function independently but integrate seamlessly through clear APIs. Our focus at Auki is particularly on collaborative perception, mapping, and positioning—critical elements that remain robot-agnostic and foundational for widespread adoption.

Why Mapping Should Be Independent from Perception

We advocate for keeping mapping separate from perception. Many current robotics companies tightly couple these two layers, which we argue is problematic because environments frequently change, requiring constant retraining of perception models if tightly linked to mapping.

Instead, we treat maps as context, external to AI perception models, enabling greater flexibility and adaptability.

Hybrid Robotics: Bridging Humans and AI

One of our innovative approaches is what we call "hybrid robotics," combining human locomotion and manipulation with AI-driven perception and mapping. This approach allows us to deliver embodied AI solutions immediately without waiting for fully capable humanoid robots.

Our recent multimillion-dollar deal proves this approach’s immediate value, demonstrating that spatial AI can deliver practical, scalable benefits today—even using just smartphones.

Future Vision: An Open Robotic Operating System

Looking forward, we aim to expand our efforts significantly. Ultimately, our ambition includes developing a robotic operating system. We see this as the future, believing the most widely used OS globally in the coming decades won't be iOS or Android, but the operating systems running robots.

We're currently collaborating closely with Mentra, who have successfully integrated our spatial awareness technology into their smart glasses (weighing only 44 grams!). This technology, coupled with wearable batteries allowing full-day use, represents a major leap toward practically integrating AI into daily human activities.

Continuing Momentum

We recently impressed major robotics companies in Shanghai with our interoperable robot orchestration system, and we're experiencing significant traction from industry leaders.

We're actively looking to expand our business development efforts. If you’re passionate about bringing innovative technology to market, reach out—we have plenty of exciting work ahead.

Join our community on Discord to participate directly in building decentralized, community-owned infrastructure for the future of AI and robotics.

Watch the whole update on X.

アウキ・ラボについて

Aukiはポーズメッシュという地球上、そしてその先の1000億の人々、デバイス、AIのための分散型機械認識ネットワークを構築しています。ポーズメッシュは、機械やAIが物理的世界を理解するために使用可能な、外部的かつ協調的な空間感覚です。

私たちの使命は、人々の相互認知能力、つまり私たちが互いに、そしてAIとともに考え、経験し、問題を解決する能力を向上させることです。人間の能力を拡大させる最も良い方法は、他者と協力することです。私たちは、意識を拡張するテクノロジーを構築し、コミュニケーションの摩擦を減らし、心の橋渡しをします。

ポーズメッシュについて

ポーズメッシュは、分散型で、ブロックチェーンベースの空間コンピューティングネットワークを動かすオープンソースのプロトコルです。

ポーズメッシュは、空間コンピューティングが協調的でプライバシーを保護する未来をもたらすよう設計されています。いかなる組織の監視能力も制限し、空間のプライベートな地図の自己所有権を奨励します。

分散化はまた、特に低レイテンシが重要な共同ARセッションにおいて、競争優位性を有します。ポスメッシュは分散化運動の次のステップであり、成長するテック大手のパワーに対抗するものです。

アウキ・ラボはポスメッシュにより、ポーズメッシュのソフトウェア・インフラの開発を託されました。

|Discord |

Keep up with Auki

ニュース、写真、イベント、ビジネスの最新情報をお届けします。

最新情報