Facing Issues with Edge Computing and Edge Impulse Studio - Any Advice?

Hi everyone,

I’m currently working on a project involving edge computing and we’re using Edge Impulse Studio for our machine learning models and edge device management. We’re encountering some issues and I’m hoping to get some advice from the community.

1. Edge Computing Challenges:

We’re experiencing the following problems with our edge computing setup:

  • Inconsistent Device Performance: The performance of our edge devices is fluctuating. Sometimes they handle tasks well, but at other times we see significant slowdowns or delays. This inconsistency is impacting our overall system reliability.
  • Connectivity Issues: We’re having trouble maintaining a stable connection between our edge devices and the central management system. This intermittent connectivity is leading to problems with data synchronization and occasional data loss.

2. Issues with Edge Impulse Studio:

Specifically, with Edge Impulse Studio, we’re encountering:

  • Integration Difficulties: We’re having trouble integrating Edge Impulse Studio with our existing infrastructure. The integration process has been more complex than anticipated, resulting in data flow issues and discrepancies.
  • Model Deployment Problems: Deploying machine learning models from Edge Impulse Studio to our edge devices has been problematic. We’re seeing issues with the performance and reliability of the models once they’re deployed, which affects their effectiveness in real-world scenarios.
  • Resource Management Challenges: We’re facing difficulties with managing resources within Edge Impulse Studio. There seems to be a mismatch between the resource allocation settings and the actual performance on the edge devices, leading to suboptimal performance.

Has anyone faced similar challenges with Edge Impulse Studio or edge computing in general? Any suggestions or best practices for addressing these issues would be greatly appreciated.

Thank you in advance for your assistance!

Welcome to the community @judywatsonn!

It sounds like you’re describing a broad set of challenges with both edge computing and Edge AI, although we dont have any specifics on hardware or use case?

Are you referring to a company or institute when you mention “we”? If you haven’t already, it would be a good idea to speak with our sales or solutions team regarding these challenges. They can provide the right guidance on the hardware you are facing these issues on.

If you can share your location and the name of your company in a private message, or directly contact our sales team, we can ensure you get routed to the correct team for further discussion. @yodaimpulse @jimbruges fyi.

https://edgeimpulse.com/contact-sales

Best

Eoin