Why IoT performance monitoring matters

IoT performance:
what’s at stake?

IoT performance (response times and errors) strongly affect the quality of service delivered by the IoT infrastructure. Let’s take an example. 

Consider an IoT platform which is collecting data from objects located in vehicles to feed public transport users with real time localization data and to optimize their route. If the end to end processing time drifts too much from real time, the value provided to these IoT users becomes worthless.

Imagine that as public transportation user you get the ETA (Estimated Time of Arrival) of your next bus only when the bus is actually arriving in front of you…. the outcome in terms of customer experience sits between nonsense and annoyance. You will stop valuing that service. 

Whether we talk about connected vehicles, smart city projects, digital healthcare, smart farming, retail or logistic applications of Internet of things, IoT analytics performance is a critical condition to generate end user value!

So IoT monitoring is an important part of the equation for IoT projects.

Challenges of IoT performance monitoring: many factors, changing conditions, lack of control on the infrastructure stacks

Ensuring the performance testing of IoT applications comes with a number of complexity factors:

  • IoT platforms are complex by nature. The number of factors impacting the delivery of an IoT service is huge: connected objects, numerous WiFi / GSM options to connect, internet path to reach the digital platform, digital platform’s location, the analytics processing time, its API responsiveness, network and application path to the devices and the client platforms.
  • Connected objects come with their limitations in terms of resources and can be hard to instrument to generate IoT performance metrics.
  • The local network used by IoT devices can vary from minute to minute from WiFi to Edge/3G/4G/5G networks with roaming impacting the performance. 
  • The network path to the edge or the cloud computing platform can also vary through time. 
  • The PaaS or IaaS computing platforms (Azure IoT Hub or AWS or GCP’s IoT platforms) are harder to instrument due to the lack of control of the IoT vendor on the technical stack (network and system layers). 

Performance in IoT is an important stake and Ops teams have a strong pressure to deliver an optimal performance in the real world and need data to achieve it. The bottom line is that IoT user experience monitoring is a must, but it requires careful thinking to know which type of performance testing should be implemented.

How you can use Kadiska for IoT performance testing

Kadiska delivers network performance monitoring between connected objects and the IoT platform

  • Visibility on IoT performance at any time: understand usage patterns (where, volume, what transactions are they using, overall performance)
  • Deep dive on performance: isolate faults immediately
    • CDN, 3rd party, cloud or edge computing platform app
    • Performance contributors
    • Errors
    • Where your users and your digital platform is performance being degraded
  • Proactive alerting powered by a global network of performance stations
  • Guided troubleshooting
  • Enhance performance
    • Benchmark your performance against your peer 
    • Get recommendations for better CDN, edge/cloud/hosting options based on where your users are located

Showing the internet path from users to application

Sounds interesting?

Find 20 minutes to discuss your requirements and discover Kadiska’s capabilities.

Let's get in touch