Neuron Industries

Industrial controls engineers today work with legacy PLCs, separate HMI packages, third-party historians, and vendor-specific IDEs, often programming in decades-old ladder logic dialects. This fragmented stack requires weeks of engineering work, per-tag historian licensing, and separate ecosystems for things like robot teach pendants. The case study describes a 1991 Siemens S5 controller running a line with 189 hardwired I/O points, illustrating how old and brittle this equipment can be. Commissioning and troubleshooting are slowed by the need to cross-reference spreadsheets, panel wiring, and disparate tools instead of a unified view.

Cortex AIC is a single industrial controller that integrates real-time control, an HMI server, a historian (Hippocampus), and the Synapse IDE served directly from the device, usable fully air-gapped from a browser. Engineers describe desired machine behavior in plain English and Synapse generates deterministic real-time Python (a scan-cycle controller.py plus an event-driven supervisor), flags unresolved edge cases and interlocks, and auto-generates unit tests that simulate the logic before deployment. It supports EtherCAT, TCP/IP-based field bus, and Modbus TCP for field I/O, claiming reach to about 99% of third-party I/O, plus open data egress via InfluxDB, MQTT, OPC UA, and ROS 2/Zenoh. The platform includes Git-based version control of code and I/O configuration on the controller, live state machine visualization with a transition flight recorder, and a Robot Studio module that treats a connected robot arm as just another device sharing the controller's program, I/O namespace, and interlocks. The case study reports an 8-hour cutover of a legacy line with 189 hardwired I/O points, with about three hours of human engineering and logic proved in simulation.