Evaluating what IoT device failure looks like in the current digital world
The universe of connected devices continues to snowball and has now surpassed the number of smartphones globally and is projected to reach 18 billion devices by 2022. Smart devices and hyperconnected systems are becoming ubiquitous, ranging from connected doorbells to smart cities to hospital systems. Organizations need to rethink how they test and monitor the devices and services they are delivering. Determining what works is very different now than in the days of static failure notices.
Simply evaluating code compliance doesn’t provide any insight into whether you are delivering the experience that a user expects. Take the smart home with a range of connected smart devices, including a security system, thermostat, sockets, oven, coffee machine and voice-enabled devices. A user expects that the devices will all be interconnected and able to react and respond to the data provided within the ecosystem. For example, as a homeowner approaches, they expect the heating system to adjust the temperature of the home to their individual preference and only open the front door once they are within five feet, and their biometrics have unlocked the door.
From a manufacturer’s perspective, merely testing what you have built no longer works. Take the smart font door example, testing if it works in isolation doesn’t suffice because you need to understand how it interacts with the entire smart home. You don’t want a scenario where when the indoor temperature increases to 80 degrees Fahrenheit, the front door opens and cools the house down.
IoT is a combination of multiple protocols, devices, operating systems, firmware, hardware and networking layers. Testing in this hyper-connected world needs to focus on the value proposition and test everything that is part of the user experience. Organizations must shift and embrace a strategy that examines the user perspective and delivers insights to optimize the experience. This requires test automation that moves beyond static failure notices to show teams where the problems lie and significantly reduce the time it takes to resolve issues. In a nutshell, testing needs to detect and fix problems in interactions of applications with people, other devices and locations, and then learn and improve the experience delivered.
To achieve this, test automation incorporating breakthrough technologies like AI, machine learning, deep learning and analytics is now essential. The pass-fail model is no longer relevant in a digital world driven by experiences. As the IoT continues to evolve, organizations need to quickly pivot and evaluate experiences, not just code compliance.
All IoT Agenda network contributors are responsible for the content and accuracy of their posts. Opinions are of the writers and do not necessarily convey the thoughts of IoT Agenda.