News · Published 22 July 2026 · Updated 26 July 2026 · 5 min read

From 24×7 robotics operations to embodied AI

Marcel Haas’s robotics context connects high-availability AutoStore operations with today’s open social-robotics and enterprise AI systems.

In one paragraph

Embodied AI is credible only when intelligence, hardware, people, and operations are engineered together. Marcel Haas’s track record spans lead engineering around high-availability AutoStore systems and the CNEXT AI Labs team’s Misty project, which connected speech, vision, automation, and live ERP data to an open robot platform.

Robotics starts with operations

Before generative AI became a board topic, warehouse automation already required software, machines, people, and service organizations to work as one system.

Marcel Haas led engineering for the technical foundation around AutoStore systems in the Swisslog environment. The public project record describes warehouse management, conveyor and picking systems, real-time computing, monitoring, automation, international on-site service, and worldwide 24×7 support.

That context matters because a physical system cannot hide weak operations behind a polished interface.

Availability is part of the design

Robotics engineering includes more than motion and perception. It includes:

  • clear operating ownership and escalation;
  • observability across software and hardware boundaries;
  • safe behavior when data, networks, or models fail;
  • incident response and service access;
  • documentation, training, and operational handover.

These are not aftercare. They are part of the system.

The Misty bridge

The CNEXT AI Labs team Marcel co-leads later developed a project on the open Misty social-robotics platform. The published record connects Azure Speech, Azure OpenAI, video analysis, Power Automate, Logic Apps, HTTP APIs, WebSockets, and live ERP data.

The project demonstrates a different robotics layer: conversation, sensing, enterprise context, and human interaction on a physical platform. It is attributed to the CNEXT AI Labs team rather than to Marcel as a solo implementation.

Toward embodied enterprise systems

The next generation of enterprise AI will increasingly leave the browser. It will reach sensors, facilities, devices, mobile machines, and personal robots.

MH – Applied AI treats that direction seriously while keeping personal robotics labeled as an emerging practice. The near-term work is readiness: architecture, data boundaries, edge inference, supervision, integration, safety, and the operating model required before autonomy can be trusted.

Sources: Swisslog project record and Misty project record.

← Back to News & Insights