Client
The client is the leading provider of wireless sensor solutions for remote monitoring. They service organizations in different industries such as manufacturing, laboratories, hospitals, and greenhouses.
Client’s Business Requirement
Our client’s product which is a Wireless Sensor Network that assists in remote monitoring, works by having a gateway connected to a business's Ethernet LAN and communicates with a smaller network using a proprietary protocol. The client utilized the services of Airbee to built a stack conforming to ZigBee specifications, test for scalability over a large network, and also test for reliability.
Details of stack requirements such as stack components, code memory, and RAM were provided by the client. The interfaces to the application layer were also provided by the client.
Airbee’s Solution
Airbee built the stack using ZigBee specifications and also extended the specifications to encompass the requirements of the client. Airbee leveraged on its testing experience, and prepared a comprehensive testing strategy that extrapolated the client’s requirements. We took into account functional as well as non-functional aspects to ensure that product is field deployable.
Using an application stub, Airbee tested the product over a larger network with a simulated condition of 32 devices. The product was tested using different topologies such as Linear, Star, and Mesh, so as to ensure better field deployability. Large volumes of data were transmitted in parallel to the gateway from multiple pods. The device was stressed to assess its performance under simulated field conditions. The simulated field conditions also involved introducing jamming signals and other peer network with high data traffic. The burn-in test method was utilized to test the performance sustainability of the devices over a period of time. The process was extended and the devices were run for 72 continuous hours and tested for failure conditions.
Observations
While testing, occurrence of defects due to asynchronous events that happen in a pattern and as well as in a random manner were tracked. In certain cases, defects surfaced after a long duration and these instances were tracked and analyzed. These defects were fixed providing us valuable insight for fine-tuning our testing process.