Akamai Technologies, the content delivery specialist, will use MWC to push what it calls its ‘Predictive Video over Cellular Capabilities’, designed for “delivering, optimising and securing online content and business applications”.

Akamai claims it’s currently in successful trials with five operators and several application providers. The idea is, effectively, to time shift video delivery to avoid congestion and bad performance, or in marketing terms “to utilise off-peak cellular capacity while at the same time better managing demand for video traffic and providing an ideal video quality experience.”

The idea is clear - shove videos across the air interface in the dead of night when data demand is low and use the actual device itself (smartphone probably) as a cache. With smartphones now hosting multiple gigabytes of main storage, the idea that you might buffer multiple videos on them doesn’t seem outlandish. The interesting bit will be the predictive part. How does that work? How will the ‘system’ know what I might want to watch when, unless it’s something episodic, I don’t know myself?

Akamai indicates that the prediction involves interrogating user preference and content recommendation engines so that the service (which is what it becomes in this context) might pre-position personalised video content to the mobile subscriber’s device.

I like that ‘pre-position’ concept. The ‘predictive’ part means the mechanism is somewhere between ‘video on demand’ and ‘broadcast channel’.  In practice it might work as a personalised catch-up TV service with a surprise or two thrown in.

An essential element in the trial solutions has been Nokia’s Liquid Applications. According to Dirk Lindemeier, Head of Liquid Applications at Nokia Networks, “combining Predictive Video over Cellular with our Liquid Applications makes it more efficient to notify upstream content providers and mobile device clients as to the most efficient time to fetch subscriber content.”

Giving MNOs the ability to sell or provide premium content services downstream to mobile subscribers and upstream to video content providers, might sound to some very much like a neutrality-busting double-dip exercise, along the lines of zero-rating. In the wake of Title II in the US and impending net neutrality legislation in Europe, it will be interesting to see how implementations of predictive video over cellular avoid business discrimination charges.