Rust in Industrial IoT: Replacing Legacy C++ in a Pilot Project For Asset Tracking Provider

We narrowed the pilot to one sensor group on a single truck using a proprietary protocol – enough to measure stability and engineering impact.

Together with the client, we mapped all integration points:

• their internal sensor, alarm, and fleet buses
• shared-memory IPC
• the C++ libraries the new module needed to call

Based on this, we designed the module’s data flow, state machine, message lifecycle, and a safe Rust–C++ FFI layer that would wrap their existing libraries.

We built the full Rust pipeline for message parsing, CRC checks, retries, and publishing data into the fleet system.

The module connected to the client’s sensor, alarm and fleet buses, and followed the same operational rules as their C++ modules. Rust’s safety guarantees helped eliminate many of the memory-related issues that previously caused instability.

Mining-truck sensors and onboard hardware weren’t always accessible, and the proprietary protocol required precise timing and edge-case handling. That’s why we built a Python simulator to reproduce real sensor traffic. It generated both valid and faulty sensor traffic to exercise the full Rust flow: parsing, retries, CRC checks, and publishing.

We tested the module under different load patterns, failure scenarios, and protocol variations. The pilot demonstrated cleaner behavior under fault conditions and reduced memory-related bugs.

One of the CTO’s goals was to understand whether adopting Rust could improve velocity of their large in-house team. Several engineers were upskilling, and our Rust pilot project proved practical for them.

Through code reviews and documentation that mapped Rust concepts to C++ workflows, we helped the team understand how to develop and maintain Rust-based modules. This equipped them to continue expanding Rust across their 70+ existing modules in future.