Is Bosch Over-Selling Video Analytics?

Published Mar 25, 2009 17:02 PM
PUBLIC - This article does not require an IPVM subscription. Feel free to share.

Analysts and pundits agree that video analytics has been significantly over-sold in the last few years. Bosch certainly did its part with its IVA Challenge [link no longer available] where tradeshow attendees tried to steal a drill positioned directly below a camera. Last year, I criticized Bosch's approach, arguing that it fundamentally mislead and missed the key challenge for video analytics.

This year, Bosch has doubled down with an even more elaborate "game" (their term, not ours). The new event [link no longer available] features:

IVA vs. You, attendees can attempt to reach the booth's football field end zone without being detected by the video content analysis technology monitoring the scene.

From a PR tactic, this is brilliant: Americans, Vegas, football and security professionals. People will love it. 

How Bosch is Over-Selling

The problem is Bosch's assertion about the game:

No more canned clips. Play around and decide for yourself whether it is ready for You.

The problem is that no trade show demo can show you that analytics are 'ready' for you.  Sell the concept? Absolutely. Bosch's approach does a better job than any canned demo of that. However, this is going to mislead people into thinking that such results can be carried into the field (I mean real deployments, not fake football fields).

Deciding on Whether Analytics Are Ready

Understanding whether analytics are ready are a function of two fundamental factors:

  • The environment you are monitoring
  • The complexity of setup
Both of these factors are tightly controlled at a trade show and unrepresentative of field deployments.  
Environment issues include vegatation, changing lighting conditions, reflection of light from the Sun, stray animals, obstructions, dirt and debris that builds over time, etc. None of this can be accurately demonstrated at a trade show.
Furthermore, the Bosch setup is going to be precisely setup by a team of experts who have tested, tuned and optimized this systems for days or weeks. This is what manufacturers do. The problem is that this part can be very difficult and a significant cost for integrators and end-users. 
By hiding these two fundamental factors, attendees will purposely be mislead into not considering the problem of false positives and the challenges of eliminating them.
At trade shows, games, booth babes and free drinks are considered good sales tactics but do they really help our customers make better decisions?