Will the decentralised approach (intelligence in camera) or server-based intelligent analytics dominate the market?
– I think there will be some kind of balance. If you run analytics on a server it can serve rather few video streams – four to six – so of course you can decentralise it but there is a drawback. Most of the environments where cameras are deployed are simply not suitable for processing the data. Not only that, but they can also be destroyed or fail, as all hardware does at some point. If you do the analytics in your server room with backup power and heat control, it’s essentially a more secure environment. At the same time, we don’t want to put too much load on the servers: the more tasks loaded on the server, the fewer the tasks that can be served, so we think a decentralised approach is better in that regard. There are both pros and cons. If you look at the software perspective, the more intelligence you put in cameras, the more often they need to be updated. The server room is a more maintenance-friendly environment, and also better security-wise. So the best compromise is somewhere
