If I only had a Brain, a Voice, a Heart - Voice Conversations & Command with Embodied AI

What's it like to control a real robot with live voice conversations? Challenges and lessons learned through provisioning a high level voice interface for my ROS2 robots. What: Ever wondered what it’s like to control a real robot with live voice conversations? See a robot in action and hear the hard-won lessons from building a high-level voice interface for ROS2 robots—what worked, what broke, and what to try next. Who should come: Robot builders, AI/LLM/ASR folks, ROS/ROS2 devs, product/practitioner types curious about embodied AI.

When & Where

Date: Tue, Oct 21, 2025 Time: 6:00–8:00 PM (2 hours) Location: Improving – Plano 5445 Legacy Dr, Suite 100 · Plano, US

Agenda

6:00–6:20 — Arrival & Networking Check-in, pizza, say hi, meet others.

6:20–6:30 — Kickoff Welcome, goals, safety notes for live demo.

6:30–7:30 — Talk + Live Robot Control • Designing voice UX for robots • ROS2 pipeline • Latency, interruptions, error recovery • Edge vs cloud trade-offs

7:30–7:50 — Q&A War stories, tooling, config tips.

7:50–8:00 — Closing & Networking Share takeaways, swap contacts.

Who Should Attend

Robotics & ROS2 builders, DevOps/ML engineers, product folks exploring embodied AI, students & mentors.

What You’ll Learn

Practical patterns for voice-controlled robots (wake words, barge-in, confirmations) How to provision and operate a high-level voice interface on ROS2 Failure modes, observability, and human-in-the-loop guardrails Bring

Questions, curiosity, and a willingness to try the mic.