Embedded firmware solution for operational efficiency

We needed physical access to prototype devices for failure analysis. As the railcars with prototypes were in use across the continent, we faced delays from a few weeks to months before receiving them. It wasn't possible to collect meaningful data after such a delay. E.g., Many times, the devices had run out of battery when we received them. In such cases, we could not tell if the failure was because of the battery running out or for any other reason. We were also looking at the issue of a few devices restarting in the middle of operations. But we did not have a mechanism to view the events just before the restart. Thus, we faced challenges in collecting the failure data.

We have contributed to many aspects of the product development and are currently engaged in several ongoing activities. We worked with the client's team on developing the firmware for monitoring parameters such as the impact on the railcar, state of the handbrake, hunting, brake pressure, and brake beam strain. We proposed and implemented a mechanism to record the system's state before a system restart. This enabled us to identify the role of the watchdog software component in the device restart issue, and we are working to resolve it. We also identified the cause of device failures as a battery issue and implemented firmware for OTA updates to the devices.

The client deployed about a hundred prototype devices with our firmware and is testing those devices. The debugger we proposed and implemented records the system state before it restarts, guiding us in analyzing one of the prototype's issues. We also helped the client identify the battery behavior as one of the causes of device failures. Our implementation of OTA updates to the devices is ready for testing. The most significant validation of our contribution is that the client continues to involve our team in several new activities within the project. We have started working on reducing power consumption and hardening the devices: Firmware reliability improvements, using crash dump, provisioning – (streamlining production and deployment process).