72. Challenges of Edge AI

4 Challenges of Edge AI

1. Vulnerable Security Feature
   Due to scattered locations, more security is required.
   

2. Limited Machine-Learning Power
   
   Edge device with low computing power (compared with the cloud) leads to under-delivery considering performance.
   

3. Model Sync
   
   Delay when updating the model running on the edge
   

4. Poor Data Quality
   
   A challenge all AI projects face