The largest growth area in data storage is sensor information. This includes areas like automotive, aerospace, industry/IoT and even security. The workflows that deal with such data are not built for scale. These data-burdened workflows require too much data to find challenges and flaws. Automation of review with AI and Machine Learning only addresses specific narrow areas of analysis. Today this means in many cases human intervened review which introduces its own problems.
If we look specifically at automotive, where Ottometric focuses as their beachhead, we see pre-autonomous systems, like ADAS struggling to achieve higher system effectiveness rates. Industry stakeholders have called out such systems as moderately effective where the risk is the end-user disabling such systems out of frustration.
The ability to better analyze where effectiveness in such systems is lacking is essential, this is an area that is lacking in current proprietary in house tools and solutions on the market today. Better simulation and better processes to emulate scenarios are essential but it doesn’t solve analysis of real generated data, specifically in automotive with the millions of miles and multi-petabytes of validation data generated from endurance testing.