Traffic and insecurity are two of the issues that most trouble residents of big cities, and Chinese companies want to gain space by helping to resolve them. At least two of them, Alibaba and Huawei, offer so-called smart cities platforms: a package of solutions that promises to use high technology to solve simple routine issues, such as a stopped car interfering with traffic.
Even in light crashes, it is common for drivers to wait for the police or other authorities to arrive, to ensure that the culprit is held accountable. While they wait, they block space. “In some cities in China, the police arrives, pulls the images from the cameras in the region and assesses whether there was anyone to blame. Thus, the innocent can leave immediately”, says Rômulo Horta, marketing director for Huawei in Brazil.
In Shenzen, a metropolis of 12 million people, there was a 15% reduction in congestion during peak hours and a 9% increase in the volume of vehicles in motion after the adoption of the technology, which includes monitoring with cameras and software capable of adjusting traffic lights and routes to optimize the use of the roads, both based on the current situation and on forecasts, made from the traffic history.
Singapore is a step further. There, cameras also monitor traffic, but when it detects an accident, the system can send drones to the scene to take images of the scene from various angles and assess whether there are any injuries. The information is passed on to the central, which decides whether to send ambulances.
The systems also aim to improve the flow of pedestrians on the streets and on public transport. Alibaba’s City Brain tool, for example, is used in cities like Hangzhou and Suzhou — with 10 million inhabitants each — to organize the entry of passengers on the subway and buses, so that travel is faster and not there are bottlenecks, such as queues to go up the escalator.
Generally speaking, the platforms are divided into four parts: data capture (cameras, motion and sound sensors), an analysis structure of what was collected, equipment for employees to receive information in real time and a fast network of internet to connect the entire system.
The hiring model, in turn, has changed. Instead of buying the equipment, you can hire it as a service and have a thousand data capture points instead of a thousand cameras, for example. The software and the fastest internet network can also be paid for on a per-use basis.
“Our suggestion is that governments start by installing the public safety and traffic part, which can be adapted to other uses”, says Ricardo Mansano, solutions manager at Huawei. The company has already taken part in the implementation of more than 160 smart cities, in more than one hundred countries and regions, including Brazil.
The city of Aparecida de Goiânia (GO), with 590 thousand inhabitants, contracted a package that includes 720 km of optical fiber and 650 cameras, in addition to its own datacenter. The idea is for faster internet to be used in other functions, such as offering free Wi-Fi in squares and improving access in public schools.
In Bahia, there are around a thousand cameras connected to a facial recognition and license plate reading system, and another 300 are expected to be added. By the end of June, the technology had identified 211 outlaws on the streets, arrested after the system, adopted in late 2018, generated an alert.
Recognition can be done even if the target is wearing a mask, and the system becomes more and more agile. An Alibaba service is capable of analyzing 16 hours of video in a minute, looking for specific information, such as checking if a suspect has passed through a certain location.
Some changes help lower costs. The cameras themselves do not need to have extra capabilities, as the image analysis will be done from a distance. You can also take advantage of records of private equipment, such as houses and businesses, which face the street.
To ensure faster internet, it is possible to create a private network to connect the equipment, the central and the employees who will use the data. “With a 5G connection, I can send high-resolution images to a police officer’s tablet on the street, instead of speaking the description over the radio,” says Mansano.
The exchange of data between different parts of the government opens up possibilities such as sending ambulance paramedics the health history of the patient they are going to fetch, catching a truck that is throwing rubble on the sidewalk or monitoring in which part of the city a student who is did not go to school.
As these platforms generate a huge amount of data, it is necessary to use several systems to analyze information and generate alerts, such as an accident or an outlaw in the crowd.
Here comes one of the sensitive points: how to ensure that governments and companies that store the information will have good control over the data? There will be detailed reports on how each person moved around the city and occurrences in which they were involved. The system could even register that someone threw the trash out of the basket on the sidewalk or forgot to indicate with the arrow when entering a street in a car.
“Many facial recognition technologies are being implemented in Brazil, but this is done without worrying about the privacy of citizens”, says Bárbara Simão, coordinator of the area of ​​privacy and surveillance at the InternetLab research center. “Many technologies still fall into a normative vacuum.”
One way of trying to balance the issue is to include in the contracts the possibility that external audits can be carried out, to ensure that the data is handled correctly. Thus, it helps to avoid cases of leaks or discrimination, such as poorly calibrated facial recognition technologies for black people, which can cause inconvenience to a citizen who just wanted to get home earlier.
Usage examples
Traffic
- By analyzing data from the cameras, the systems detect problems, such as an accident, and send an alert to other drivers to deviate from the area, in order to avoid congestion
- In the event of a crash, an image file can show who was at fault, without having to wait for expertise, which frees up the path more quickly
- Adjust the timing of traffic lights to improve fluidity, based on average traffic per day and time and real-time data
- Collecting urban tolls using automatic plate reading. You can also fine cars that are circulating without up-to-date documents
Public transport
- Sensors can detect which trains or buses are too full and thus reinforce the fleet more quickly
- Facial recognition can be used to charge the ticket, which does not require the use of tickets.
Public security
- Face-recognition cameras let you scan crowds and find outlaws
- Sensors can detect the sound of gunfire, explosions or smoke, speeding up delivery of help
urban occupation
- Software can analyze images to detect irregular occupations, such as invasion of environmental areas, constructions or opening of businesses without a license
the parts of the system
fundraising Cameras, motion sensors, heat sensors, etc., to generate real-time street data
Management Software capable of analyzing large amounts of data and generating alerts
Delivery of information Control rooms to monitor data and equipment to send information to agents on the street
Networking Fast, sometimes private, internet connections to connect the entire system
.
