One of the recent innovations proposed in video surveillance is the introduction of blockchain technology to reduce expenses and improve efficiency. We recently spoke to Robert Pothier, CEO of Faceter, a company that is pioneering this effort.
One of the recent innovations proposed in video surveillance is the introduction of blockchain technology to reduce expenses and improve efficiency. We recently spoke to Robert Pothier, CEO of Faceter, a company that is pioneering this effort.
So how exactly can blockchain be used in video surveillance? According to Pothier, the technology could be used for two things:
1. Building a decentralized infrastructure (so-called fog computing), which will meet the requirements of the facial recognition tasks, and
2. Receiving payments from customers, fixing orders in smart contracts and transferring payment to miners (participants of the company’s fog network, with whom it will sign up smart contracts).
“We need a decentralized infrastructure because it will help us to provide a consumer version of our solution and make our product more affordable for homeowners and small businesses globally,” he said. “Today, facial recognition and computer vision technologies are costly and resource-intensive. We have a working solution, and we have an interest in potential business users.”
However, current classic infrastructure architecture can't be used for providing the mass market solution – it is too costly. So, Faceter plans to develop its solution to be able to pass the processing work over to existing miners in the blockchain industry. This decentralization could drastically reduce the price of the solution.
“Even if we offer Ethereum miners five times more than they are currently getting to mine cryptos, it will still be financially feasible for us,” Pothier said. “With PoS becoming more popular, there is a growing concern for miners as they have the hardware but mining less as mining pools continue to get larger and more demanding.”
In the future, smart contracts with customers can also become part of the collective security network, integrated with police databases. Thus, the police will be able to distribute pictures of wanted criminals, lost kids, stolen cars, and so on.
Advantages of using blockchain in video surveillance
The appeal of blockchain is the price and efficiency factors, which are after all the two main concerns that customers all over the world share. From an installer’s perspective, this could be an advantage as more and more clients seek cost-efficient solutions now.
“As I mentioned above, the use of blockchain technology can significantly reduce the cost of face recognition systems and improve the efficiency (speed/quality) of the video stream processing,” Pothier said. “I suppose that all the participants in the video surveillance market will benefit from our technology – both private and corporate users, as well as hardware manufacturers, installers, etc. Users will have the opportunity to make their lives safer and more comfortable, while businesses will have more work since video surveillance systems will be more common.”
He further added that every retail owner can contribute and earn profits using Faceter technology and, of course, the blockchain, because blockchain is extremely convenient for data exchange and maintaining shared databases.
“So, if you are a business owner, then by installing a video surveillance system with our software, you will be able to collect and share marketing data, have the complete database of your customers and their needs and preferences, and also help in capturing criminals and searching people through a very quick online database comparison,” Pothier said.
Are there any challenges?
No technology is without challenges. According to Pothier, critical among them include:
1. Minimum delays in video processing
Every second of video processing represents some stages – from splitting into frames to detection and vectorization of faces.
“In the current version of our product, all of these stages are passing through only one computer,” Pothier said. “But in the distributed version, we will separate all tasks between different nodes so that no one has a complete source image. Obviously, this separation will add overhead, but we are looking for ways to reduce it.”
2. Fault tolerance
Making many calculations, every computer can be unstable. It may lose access to the internet or just shut down.
"We have to configure our system in such a way that it can automatically detect such failures and redirect the tasks to other nodes,” he continued. “Since we have not found any ready-to-use solutions in similar systems (SONM, Golem, and others), we will develop such a system by ourselves.”
3. User data security issues
Pothier added that since the launch of the Faceter project, very important changes in the protection of user data have been taking place. These include regulations like the GDPR, which puts forward quite severe demands for this matter.
“This extraterritorial law cannot be bypassed, and we will fully observe it,” he said. “So, this is another important challenge for our team.”