Expanding Dyad for embedded engineering
JuliaHub appears to be broadening its Dyad platform toward a comprehensive, safety‑critical, embedded‑engineering work‑space, adding desktop distribution, AI‑driven modeling, and traceable code generation for edge hardware.
Public signals include a series of product releases and feature announcements that emphasize hardware‑in‑the‑loop workflows, support for microcontroller and Linux targets, aerospace‑focused use cases, and enhanced security and tooling for regulated environments.
-
The source shows JuliaHub launched Dyad 3.5 with support for modeling discrete controllers together with physics, combined simulation, and generation of C code for embedded hardware, including model predictive control, extended PID tuning, and AI Agent sessions that can keep packages loaded between compilations.
-
The source shows JuliaHub highlighted Dyad for hardware-in-the-loop workflows, traceable C‑code export for safety‑critical use cases, and flexible deployment to microcontrollers or Linux targets such as Raspberry Pi.
-
The source shows JuliaHub launched Dyad 3.4 as a standalone desktop app with built‑in Julia and Git tools, as well as multi‑agent AI with parallel sub‑agents, pointing to a more independent engineering work‑space rather than solely a cloud‑based development experience.
-
The source shows JuliaHub positioned Dyad 3.0 toward aerospace teams and use cases such as flight controls, aerodynamics, cabin systems, asset monitoring, and repetitive modeling/testing.
-
The source shows JuliaHub highlighted a Dyad demo where a rotary pendulum controller is taken from a multibody model and closed‑loop simulation to automatically generated C code that runs on a Raspberry Pi with one click, reinforcing the focus on embedded deployment and code generation.