Traditional cloud computing models sends data from the device to the cloud for data analysis and the decision is sent back to the device for implementation. The agility of cloud computing is great but not enough to overcome certain challenges such as latency, bandwidth, processing the data for real-time decision making, costs associated with data transfer between cloud and edge. Cloud AI models often needed to be trained with data collected from devices, making it difficult and time consuming to apply AI and generate insights. AI with edge computing will solve the challenges faced in cloud, as the inference and training is totally moved towards the devices.
In brief, this research provides the following:
- A brief snapshot of convergence of edge computing with AI
- The challenges of existing cloud AI models and how edge can solve
- Key participants delivering intelligent edge AI solutions for different industries
- Highlights of innovative future applications through convergence models
- Roadmap and key milestones to achieve in the near, medium and long term to make devices, machines and things more intelligent.