Transport security: From a measurement to an entire network

Transport security

Share this content

Facebook
Twitter
LinkedIn

Alex Osypov, CEO & Founder, FF Group discusses transport security market trends and the value of real time data.

How has the transport security market changed in Europe?

It is fair to say that the European market is probably the most mature from a road and public transportation network perspective.

Transport flows have undoubtedly increased over the years and the aspect of the market that we are involved in is getting equally mature in terms of measurement.  

Cameras are not that expensive anymore, but they are more powerful.

Therefore, using cameras as data sensors works well and provides a possibility to companies to measure data in a different way.

With different stakeholders appearing, the main shift in the market is that there are now more data flows and more possibilities to measure data.

This also means that more businesses are working to decide what to do with their data.

In the past, each sensor was like a point of measurement, with the main stakeholder typically the police.

But gradually, requests are shifting towards municipalities, which need to measure and plan traffic flows.

As a result, I’d say we have shifted from a point of measurement to an entire network.

Instead of having one specific point, it’s now become a network of IoT connectivity sensors capable of taking measurements. 

The CCTV related transportation market has shifted from a point of measurement to a network of measurement, with the density of data increasing drastically.

Almost overnight, we went from having a stream of data to having a sea of data – data lakes, data warehouses.

The reason for this major change is related to transportation.

We are aware of many changes because of our partnerships with the likes of Axis, Milestone and Hanwha.

Can you tell us about your work in transport security please?

We have two major focuses as a business.

Our primary focus years ago was related to sensors, such as our EDGE systems and purely producing data at the edge.

Though we are still doing that, given the recent shift, we feel that we should be responsible for not only generating metadata, but processing it – this is where FF DataFactory is extremely valuable.

Our work is now on both sides of the data flow.

One focus of the business is on the generation of data, and, on that part, we are being given more powerful processors and cameras.

With this technology, we can produce more and use deep learning to gather rich metadata from a single sensor.

On the other side, we are involved in making that huge amount of data available for customers.

For example, when you have ten small parking lots which cars come in and out of each day, there is a lot of data being generated.

Now imagine there are thousands or potentially millions of cars a day.

You need special tools to handle this data.

FF Group is deeply involved in producing complex insights but easy to use tools for customers looking to process and manage data generated by businesses and cities. 

How valuable is real time data from a traffic management perspective?

Good question. My answer relates to the first point where I discussed the shift we are witnessing.

We used to work in data forensics, meaning that we collected variable data and we conducted research into what happened.

If you wanted to look at how many cars passed through the city last year, the police would have been the key stakeholder in that research and investigation. 

In a real time investigation, however, if you have multiple cars traveling per second, driving in multiple lanes and being picked up by multiple road cameras, there is more data than you can understand.

If you need to quickly find information out about a particular car, information needs to be processed and sent back immediately.

Police officers need to react immediately.

It’s a huge amount of data to process instantly but, given the technical possibilities and technologies which are emerging, whilst it’s not happened yet, we are on the verge of being capable of processing mass data in real time. 

It’s a very fast measurement and that’s why edge processing is so important. In the cloud or on the AI data analytics platform, you have less time for processing. You need to process raw information on camera – measure it, understand it.

From there, it can be sent to the cloud or AI platform for high level logic processing before the insights are gathered and the information is sent back to the point of action – in this case, the police officer. 

From measurement to action, it should take seconds. Real time data is incredibly valuable and it will open up new possibilities.

A lot of technology will appear over the next few years and we’re going to see that real time environments apply not only to forensic research, but for processing huge amounts of data by utilising edge sensors.

Can you tell us more about your LPR and MMR technology?

LPR and MMR technology is a mature field of technology that started over ten years ago – who would have imagined that it would provide so many opportunities for R&D and production for a company like ours?

LPR and MMR are successfully used and are in high demand in the market.

Now, however, the areas of application need to be expanded.

New AI cameras expand the possibilities of obtaining data and are already finding customers in the market.

To adapt AI for the edge, FF Group uses its achievements in synthetics to improve algorithms and train neural networks; we will translate AI technology into a commercial product. 

It is important to create your own architecture, your own methods of training networks.

Whoever is faster and better in all this, achieves success in AI.

Our next step is to train our technology in real time.

All this will influence getting accurate data quickly to make the right decisions and manage big traffic flow data with smart platforms.

On the edge side, we are providing more metadata; two years ago, we couldn’t support license plates of low quality.

Now, we can widen that spot, regardless of whether there are poor weather conditions or bad images; everything can now be processed, meaning we can provide more with the same sensors. 

Customers can get more information and control.

AI capabilities allow us to be able to measure a certain type of car, separate ambulances from fire engines and so on.

Again, this is all happening at the edge.  

We provide customisable software, not just from a display perspective, but from a processing perspective.

You can create your own software to process your own data using an AI road data platform (such as FF DataFactory) because, at the end of the day, only owners of the business or operators of the business have an insight which should be calculated – we are just the technology producers.

We are building tools which allow customers to get involved.

1-ISJ- Transport security: From a measurement to an entire network
Alex Osypov, CEO & Founder, FF Group
Newsletter
Receive the latest breaking news straight to your inbox