Yup, not only have I been finding ways of controlling the higher functions of robots through a clustered management system housed on the cloud, I have also taken it upon my self to give it an API. Next up, the ‘Kill All Humans’ subroutine…
How it got started:
So the truth behind this system and it’s original purpose was the health automation of large clustered systems. It seemed like if I took it a step further I could use it for control and instrumentation. From there, it really is just a stones throw to clustering that control and instrumentation into a sort of cloud-fueled hive mind.
One of the really neat aspects has been the resource API for sending feeds and data streams up to an analytical engine that returns multidimensional state maps. These state maps often consisting of a heat-map denoting cluster peers vs anomalous behavior vs given epoch allow for complex behavioral data to be generated. The system for creating these maps is provided by a local Vancouver company by the name of Metafor Software.
Metafor Software is run by some friends of mine and focuses on providing anomaly and drift detection services for cloud, virtual, and physical infrastructure. In their words “Automated detection of unexpected changes and anomalies in your environment” They also have a wonderful alpha product which crunches data streams into near-realtime cluster-maps.
Behavioral disorders and dissociation:
So the idea then becomes that this behavioral data we create is mapped into a cluster-pattern by approximating the social norm of the clusters behavior based on the weighted mapping of each peers behavioral chart. This will allow us to understand the desires of the cluster as a cohesive unit, despite a few systems potentially acting abnormal. The value in this is to be able to enforce a behavioral impression on the outliers like a hive mind.
At this point we have this nice behavioral ecosystem where we can not only observe the organic behavior that is created by the cluster through it’s activities, but also enforce a status quo amongst all the peers.