A Human City Technology
Since most people in the world have moved into cities, the question of arranging life in urban space rationally is becoming more and more relevant. Widespread introduction of the ‘smart city᾿ has manifested a megatrend – to a certain degree involving modern megacities – from New York, Tokyo and Singapore to Shanghai, Delhi and Moscow. In fact, a ‘smart city’ is generally a human-orientated, convenient, secure and environmentally friendly environment based on modern energy, transport and informatics solutions. This also includes security and video surveillance systems built on neural networks and artificial intelligence. Russia is among the leaders in introducing these technologies into everyday urban life.
The 2018 FIFA World Cup is NtechLab among the clearest manifestations. During the championship, among other things, computer vision technologies developed by NtechLab, were widely tested and demonstrated excellent results. e deployed system connected to more than 500 city surveillance cameras allowed more than 100 individuals included in offender databases to be detained during the championship. In addition, they helped in retrieving the Budweiser sponsorship cup stolen during one of the matches. This is just one particular case when video analytics systems were used.
Even today, NtechLab’s algorithms allow dozens of criminals to be identified at transport facilities every month in one Latin American capital; they help regulate people streams and avoid crowds at train stations in India; to build a smart and secure 21st century digital city in one of the largest Gulf countries. Computer vision from NtechLab provides antitheft protection in Russia’s largest retail stores, more than 30,000 monthly pre-trip inspections with instant authentication, and secure online banking.
In fact, the capabilities offered by computer vision algorithms go beyond all this. Based on experience, we can say that, on the whole, it makes the city safer, more convenient and environmentally friendly. A real-life task would be to adapt thousands of cameras in an urban metropolis to notify the ambulance service when someone falls down in the street or of an attack of an illness (a stroke, for instance). Along with vehicle traffic management, car and public transport recognition opens up broad prospects for creating an infrastructure orientated on people’s real needs. In the ‘smart city’ of the future, you will not have to wait for a bus for a long time and then push inside, as depersonalised passenger data routes and time intervals will make for more accurate planning.
The city of the future is a queue-less and crowd-less one. From the perspective of the coronavirus pandemic, this becomes especially relevant. When introduced, such systems will enable efficient resolution of another acute problem in modern metropolises – rubbish and its timely removal. Each removal of rubbish entails costs, while the bins by condominia are filled unevenly. For maximum efficiency, the vehicle routes need to be planned to avoid ‘empty’ points along the way.
Standard measures such as urban video surveillance solve this problem only partially, whereas manual checks on images from thousands of cameras and routing development require extra resources. Smart video analytics can assist by analysing both the fill level of the rubbish bins and the quality of discharge work, including third party contractors. The best routes for rubbish lorries can be generated automatically from data from cameras, cutting fuel costs and increasing staff efficiency. We have to admit that most people still mistrust AI. The stereotypical mind shaped largely by the media and the hyped dystopias create a picture of some robotic mind empowered to make independent decisions, determining who should do what.
In practice, AI is nothing more than a human decision support system. As in security or any other industry, the system only suggests what to focus on particularly, leaving the final decision with the human operator. The robot can eliminate routine operations and processes that can be difficult to control. So, the objective is to help, not control and supervise. On the whole, as AI advocates, we can declare with confidence that robots will be capable, in the very near future, of taking on operations humans simply cannot cope with owing to natural limitations. They will look through huge bulks of video, analyse gigabytes of data, react instantly to dangerous situations such as a fight, a heart attack, or child abuse. Algorithms will assist in locating people who are lost and cannot help themselves.
No matter how paradoxical it might sound, computer vision will make the city of the future more human. In turn, NtechLab intends to develop and actively promote such technological solutions not just in Russia but all over the world. Today, the company has more than 30 big clients in around 15 countries in various regions, including the CIS, Europe, the Middle East, Latin America and Southeast Asia. NtechLab products are used in public and corporate security, retail, finance, entertainment and hospitality industries. Every year, the company is top ranked in international computer vision competitions.
The latest achievements include prizes in international competitions for recognition of actions and the so-called ‘deepfakes’, or fake faces, by video. The company is working continuously working to boost the accuracy and speed of algorithms, expanding the domains where computer vision may assist humans. Last year, when the world faced the coronavirus pandemic, we presented our dedicated anti-covid solution. It helps prevent crowds in public places, ensure that people maintain social distancing and wear masks.
The solution was in great demand both in the regions of Russia and abroad. We are continuing our active development. Last autumn, NtechLab raised over 1 billion rubles from the international consortium of the Russian Direct Investment Fund (RDIF) and some leading Middle East sovereign wealth funds. This was not only recognition of the authority of NtechLab and the high quality of its solutions, but also a solid foundation for further technological expansion to the benefit of people and humanity in general, in cities and beyond.