A new remedy for IoT device quality assurance
Code is rarely perfect at the first release. It is very common with remaining issues like missed bugs, occasional delays and performance issues, or features that are difficult for the users to understand. With more and more devices having IoT cloud connectivity, it becomes possible for developers to learn how their software behaves during real-world usage. This allows for data-driven product improvements and for quickly revealing remaining issues. The earlier you can fix them, the fewer customers will be affected. The updated software can be distributed as a secure over-the-air (OTA) update using services from leading IoT/Cloud vendors.
Percepio DevAlert is a cloud-based monitoring service that makes it easy for device software developers to report any condition of interest, such as an error, warning or a timeout condition. This can be used both in system testing and in customer deployment and provides automatic real-world feedback back to the developers within seconds after the very first occurrence. The provided diagnostics includes a compact software trace for Percepio Tracealyzer, containing the timeline of software events just before the issue was reported. This gives the context of the issue, often crucial for understanding and fixing the problem.
This approach creates a feedback loop that allows for rapid continuous improvement, which is a cornerstone of the popular DevOps methodology. The monitoring provided by DevAlert allows for fully embracing DevOps in embedded/IoT software development. This way, you can rapidly improve your software after the initial release and make a great product that beats the competition.
Without automatic feedback from deployment, you rely on the end users to report any issues. It may take considerable time before such a report reaches the developers. Moreover, you also rely on the customers to provide sufficiently detailed information to allow for identifying and solving the problem. A vague error report like “it crashes several times per day” may require weeks of guesswork to find a likely cause, and even then, you don’t know if you really solved the right problem. Months or even years may pass before the problem is finally fixed, if ever, during which many customers are affected. Imagine how much customer frustration (and debugging time) that can be saved by automatic, instant notifications directly to the developer team, on the very first occurrence of the issue, and with detailed diagnostics that shows what happened.