More use cases for Edge AI

Since my team and I started learning about Edge AI, we always find interesting ideas and use cases. Now that we’re using it to train our vision AI model, I challenge myself to read and learn more about it.

I found that moving the collection, storage, and analysis of data from the cloud to the source not only produces almost instantaneous responses, but it also heightens security. As Edge AI keeps data in the local system, it eliminates the risk of exposing otherwise private data to the public on the cloud. With data power in proximity, it is said to bring better results.

Other than uncovering these values, we found more practical application of Edge AI, such as:

• Autonomous Vehicles - Self-driving cars have been around for some time now, and it’s largely dependent on cloud data to formulate actions. However, with Automotive Edge, there is less need for connectivity and instead, more autonomy for vehicles to respond to its surroundings. The microdata center within the car enables faster, more accurate data processing and reaction. This model can potentially save lives and improve road safety. (i.e. detecting pedestrian to prompt an immediate halt)
• Security Surveillance and Monitoring - With machine learning-enabled local servers, connected cameras are able to recognise people and detect movements and suspicious activities in real time. Set parameters will trigger action of other connected devices, for instance, a smart alarm system or calls the attention of responsible party to respond. This application of Edge AI improves security and safety in banks, other businesses or properties.
• Body Monitoring - Wearable technology, like smart watches, are able to collect heart rate, sleep and physical activity data, and among others. The information can be used in predicting stress levels or possible health conditions prompting the wearer to make lifestyle changes or see a doctor to improve health and wellness.
• Industrial Maintenance - While most factories are already automated, it can be taken to another level further. Edge sensors attached to machines are able to sense vibration, temperature or noise that the local data center can prompt machine inconsistency or failure and predictive maintenance before it becomes critical to the operations.

We are discovering new use cases for Edge AI every week and our research continues - we keep asking this question every week: What we can learn next week and how fast can we run an experiment?

Onwards, we will continue our journey to explore new actionable insights to find wildly important problems that Edge AI can help solve.