The State of Video Analytics

The State of Video Analytics 

First of all, what do we mean by video analytics? The term “video analytics” is an overarching term encompassing several solutions – most frequently, it is used in relation to Artificial Intelligence (AI) and deep-learning. The results IFSEC Global’s recent survey suggest that businesses are becoming increasingly convinced of the return on investment and the benefits that video analytics can bring to their operations. For instance, 71% of respondents feel that video analytics can provide value, 24% wish to evaluate it further and only 5% of respondents don’t believe that video analytics could provide real value. When evaluating the survey data, it is worth to point out that of those who were ‘unsure’, over 70% were not actively using AI or deep learning analytics already. This would suggest that these respondents’ initial answers came from a sense of caution. 

IFSEC Global Survey

Through IFSEC’s research, they have found that since 2019, there has been a significant growth in the value recognition of video analytics. Furthermore, 54% of the market is not currently actively using AI/deep learning analytics in their surveillance systems, so it is clear that there is a lot of opportunity in the market. 

IFSEC asked respondents to list the 3 major barriers to implementing AI/deep-learning (key growth areas in analytics) in their surveillance systems (or, for those that have implemented them, what were the initial barriers to upgrading).

85% of respondents cited cost as one of their top 3 challenges and 50% also included ‘concern over ROI’, further expressing a financial concern. Evidently, the business size did not seem to affect the outcome of this answer, with the proportion of small and large companies citing financial concerns as relatively similar. As technology continues to develop, however, affordability of solutions will likely increase for the majority of the market. 

GDPR/data concerns were the next major barrier to implementing AI/deep learning, directly after ‘cost’. 52% of respondents placed this in their top 3 obstacles, highlighting how seriously companies are taking data protection. It is worth noting for such concerns that data protection assessments, cyber security audits and strict data retention procedures can all
help to mitigate the risks. 

It is also noteworthy that 46% of respondents cited ‘upskilling employees’ as one of their top 3 obstacles. The engineering skills gap in the installation and integration sector is well publicised. This perhaps demonstrates that investment in operator and mid-level management skills are necessary across the sector. It’s important that those on the frontline of the surveillance operation learn how to best utilise the video analytics available to them. Amongst the ‘other’ responses, which were left open for comment, ‘educating the customer’ and ‘concerns over accuracy’ were chosen as the more common barriers. 

What are the Benefits of Video Analytics?

As part of their report, IFSEC also explored which type of technology that respondents view to be the most beneficial to their business. In their survey, IFSEC attempted to cover some of the most well-known and popular types of video analytics (reducing false alarms, loitering detection, facial recognition, customer behavioural analysis, detecting slips, trips and falls, general footfall analysis, and directional footfall) and asked respondents to rate them in order of importance from 1 to 7. 

46% of respondents ranked “reducing false alarms” as their top reason for investing in video analytics. This was ranked far ahead of the other options, so it is evident that finding methods to reduce false alarms are a major priority. Next was ‘loitering detection’ and ‘customer behavioural analysis’ which both scored relatively well. Such solutions that identify patterns of behaviour, such as suspicious activity, are becoming more popular and trusted amongst operators.

Along with the other types of video analytics, software is now able to detect liquid spills. SpillDetect is our latest solution that uses AI and machine learning to automatically detect liquid spills. It uses existing CCTV hardware and alerts staff as soon as a spill has been detected. 

No longer should surveillance systems simply be viewed as just a compulsory tool for loss prevention, but rather as a set of business intelligence tools that can increase profitability, protect profits, and deliver rapid ROI.