> Markdown version of [/videos/100230-robots-among-us-advances-in-ai-for-everyday-androids](https://www.wearedevelopers.com/videos/100230-robots-among-us-advances-in-ai-for-everyday-androids). Every page supports `.md` or `Accept: text/markdown`. Links point to the HTML versions so they work for humans too. Agent guide: [/agents.md](https://www.wearedevelopers.com/agents.md).

---

# Robots Among us: Advances in AI for Everyday Androids

Rigid, hardcoded hardware no longer limits robotics. Today, open-source Robotic Foundation Models let developers train affordable everyday androids using natural language and VR.

- **Speakers:** [Florian Gather](https://www.wearedevelopers.com/@florian-gather), [Thomas Endres](https://www.wearedevelopers.com/@thomas-endres)
- **Event:** World Congress 2026 Europe
- **Published:** July 10, 2026
- **Duration:** 29:51
- **URL:** https://www.wearedevelopers.com/videos/100230-robots-among-us-advances-in-ai-for-everyday-androids

## Summary

The narrative traces the evolution of robotics from early mechanical constraints to today's agile, AI-enabled humanoids like the Unitree G1. Historically, deploying a functional android required solving steep hardware challenges, including prohibitive actuator costs and spatial navigation using LiDAR and Simultaneous Localization and Mapping (SLAM). Early approaches to movement relied heavily on hardcoded inverse kinematics solvers, which frequently resulted in rigid, catastrophic failures during dynamic real-world execution.

To overcome these physical hurdles, reinforcement learning initially bridged the gap between simulation and the physical world. By virtually rewarding balance and punishing falls, humanoids can natively learn fluid mobility before taking a physical step. However, because reinforcement learning demands complex, custom simulations for every single manual task, the industry is accelerating toward Robotic Foundation Models (RFMs). Conceptually functioning as the robotics equivalent of Large Language Models, RFMs combine vision-language models with diffusion transformers. This architecture maps environmental understanding and natural language prompts directly into physical action decoders, bypassing the need for rigid step-by-step programming.

This shift fundamentally lowers research and deployment barriers by reducing the reliance on hyper-expensive, proprietary hardware. Using open-source frameworks, developers can construct inexpensive 3D-printable robotic arms and collect high-quality training data through rudimentary teleoperation via VR headsets. With minimal datasets, these models can be fine-tuned to execute nuanced domestic tasks like folding laundry or sorting objects. The ultimate trajectory of the field points toward adaptive motion optimization, where androids synthesize capabilities directly from internet videos to autonomously navigate and assist in human environments.

**Keywords:** robotic foundation models, humanoid robots, simultaneous localization and mapping, inverse kinematics, reinforcement learning simulation, vision-language-action models, diffusion transformers, unitree g1, open-source robotics hardware, teleoperation data collection, robotic actuators, lidar spatial navigation, adaptive motion optimization, 3d-printable robotic arms

## Chapters

1. **Evolution of humanoid robots and automated machines** (01:19) — The history of robotics traces the transition from ancient mechanical concepts to prompt-following machines.
1. **Core hardware and software components of robotics** (06:54) — Modern robots rely on complex actuators, lidar sensors for spatial awareness, simultaneous localization and mapping, and inverse kinematics solvers.
1. **Training robot locomotion using reinforcement learning methods** (15:00) — Reinforcement learning inside simulation environments allows robots to organically develop stable walking patterns before transitioning to physical environments.
1. **Understanding the architecture of robotic foundation models** (18:57) — Robotic foundation models combine vision-language models with diffusion transformers to translate instructions into mechanical actions across separate platforms.
1. **Training custom foundation models with affordable robotic arms** (22:28) — Collecting teleoperation data with inexpensive 3D-printed physical hardware enables foundation models to master practical tasks like sorting items and folding clothes.
1. **Future directions in prompt-based robotics and video imitation** (26:21) — Future robotic foundation models aim to execute complex plain-language prompts and learn new functional tasks directly by imitating internet videos.

## Related Moments

- [Transitioning from rigid machines to autonomous AI robots](https://www.wearedevelopers.com/videos/472-enhancing-ai-based-robotics-with-simulation-workflows) (from "Enhancing AI-based Robotics with Simulation Workflows")
- [Expanding robotics beyond humanoids across diverse industries](https://www.wearedevelopers.com/videos/100039-physical-ai-for-the-next-wave-of-industrial-digitalisation) (from "Physical AI for the Next Wave of Industrial Digitalisation")
- [Shifting focus from large language models to physical AI](https://www.wearedevelopers.com/videos/100039-physical-ai-for-the-next-wave-of-industrial-digitalisation) (from "Physical AI for the Next Wave of Industrial Digitalisation")
- [Future predictions for autonomous agent environments and robotics](https://www.wearedevelopers.com/videos/1629-beyond-chatbots-how-to-build-agentic-ai-systems) (from "Beyond Chatbots: How to build Agentic AI systems")
- [Progressing from standard robotics to cognitive learning systems](https://www.wearedevelopers.com/videos/86-rpa-in-the-public-sector) (from "RPA in the Public Sector")
- [Deploying inside-out foundational models onto mobile arms and humanoids](https://www.wearedevelopers.com/videos/100039-physical-ai-for-the-next-wave-of-industrial-digitalisation) (from "Physical AI for the Next Wave of Industrial Digitalisation")

## Related Articles

- [WWC24 Talk - Scott Hanselman - AI: Superhero or Supervillain?](https://www.wearedevelopers.com/magazine/469-wwc24-talk-scott-hanselman-ai-superhero-or-supervillain)
- [Stephan Gillich - Bringing AI Everywhere](https://www.wearedevelopers.com/magazine/489-stephan-gillich-bringing-ai-everywhere)
- [MLOps And AI Driven Development](https://www.wearedevelopers.com/magazine/82-mlops-and-ai-driven-development)
- [How we Build The Software of Tomorrow](https://www.wearedevelopers.com/magazine/120-how-we-build-the-software-of-tomorrow)

## Related Jobs

- [AI Software Engineer (Germany)](https://www.wearedevelopers.com/jobs/48317-ai-software-engineer-germany) at **Sunhat**
- [Machine Learning Engineer](https://www.wearedevelopers.com/jobs/ext/588393-machine-learning-engineer) at **Twilio**
- [AI Operations Manager (all genders)](https://www.wearedevelopers.com/jobs/48263-ai-operations-manager-all-genders) at **envelio**
- [Senior AI Agent Software Engineer (Go, Python) (m/f/x)](https://www.wearedevelopers.com/jobs/48277-senior-ai-agent-software-engineer-go-python-m-f-x) at **Dynatrace**
- [Machine Learning Engineer](https://www.wearedevelopers.com/jobs/ext/1355348-machine-learning-engineer) at **TWILIO**
- [Staff, Machine Learning Engineer (L4)](https://www.wearedevelopers.com/jobs/ext/1202639-staff-machine-learning-engineer-l4) at **Twilio**