China's Golden Shield

October 2001

Greg Walton

« previous | 5 | 6 | 7 | 8 | 9 | 10 | 11 | 12 | 13 | 14 | 15 | 16 | 17 | 18 | next »

A Network That Knows Who and Where You Are & Box 3: "Neutral" Technology at Tiananmen Square


Right to Association

Everyone has the right to freedom of peaceful assembly and association

– The Universal Declaration of Human Rights. Article 20.

One of the stated objectives of the Golden Shield project is the establishment of a nationwide network of closed-circuit television or CCTV cameras in public spaces to improve police response times to outbreaks of social unrest. (57) The incredible range of surveillance cameras on display at Security China 2000 indicates the extent to which technology is more complex today than years ago. New circuits allow the camera to ignore bright, light-emitting objects within its fields of view; miniaturization allows easier concealment; infrared cameras allow surveillance in darkness.

"The China CCTV market is worth US$350 to US$400 million per year," estimated Gerrit Hurenkamp, Development Manager for the US-based Pelco International. "It’s a good market but difficult to get into. They are well educated here and know what they want. You can’t just dump any product."

As video surveillance electronics become increasingly sophisticated ever-greater bandwidth is required to transport the stream of images produced from remote locations to control rooms. Such a system requires advanced network architecture, capable of spanning a country as large as China, and Nortel’s presentation of its JungleMUX system at Security China 2000 spoke directly to that need. Closed-circuit video signals in Nortel’s JungleMUX network are transported over a wide area network (WAN) operating between 1.6 to 44 Mbps and accessible by all the nodes in the network. Each video source (camera, VCR, etc.) is digitized using a user-configurable compression algorithm. This provides an efficient and scalable CCTV transport solution.


Box 3:
"Neutral" Technology at Tiananmen Square

Following the Tiananmen Square massacre in 1989, the Chinese authorities tortured and interrogated thousands of people in an attempt to identify the demonstration’s organizers. But even if the students and workers had resisted the terrors of the secret police, the hapless demonstrators stood little chance of anonymity. Stationed throughout Tiananmen Square is a network of UK manufactured surveillance cameras, designed to monitor traffic flows and regulate congestion. These cameras recorded everything that transpired in the months leading up to the tanks rolling into the square.

In the days that followed, these images were repeatedly broadcast over Chinese state television. Virtually all the transgressors were identified in this way. Siemens Plessey, which manufactured and exported the cameras, and the World Bank, who paid for their installation, claim they never had any idea that their "technologically neutral" equipment would be used in this way. However, in 1995 the World Bank authorized the funds to set up the same traffic flow system in Lhasa, the capital of the Tibet Autonomous Region. Lhasa is not, as yet, known for having problems with traffic congestion; besides, the area in which the traffic flow system is in operation is solely for pedestrians. (56)

Recording can be in several modes: real-time, of varying qualities, and time-lapse. Real-time is like regular TV (at 30 frames per second, showing full motion). Time-lapse selects only a few frames per time period to record. The main advantage of time-lapse is that it allows one tape to record for a much longer time than real-time recording, a particularly useful feature for archival purposes: high-resolution 700x480 colour pictures at 2 frames/sec, using about 400 kbps per camera. The flexibility in bandwidth allocation of JungleMUX Video Mapper allows for requesting higher resolution images and more frames per second from a specific camera, at any given time (up to broadcast-quality 700x480, 30 frames/sec NTSC colour signals, using about 6 Mbps [MPEG-2 quality]). The system even allows for ambient audio channels for public surveillance applications.

The revolution in urban surveillance will reach another level of control altogether – once reliable face recognition software becomes the norm. It will initially be introduced at stationary locations, like turnstiles, customs points, security gateways, etc., to enable a standard full-face recognition to take place. We are at the beginning of a revolution in "algorithmic surveillance" – effectively applying artificial intelligence routines to data analysis via complex algorithms, which enable automatic recognition and tracking. Such automation not only widens the surveillance net, it narrows the mesh. (58)

One company at the forefront of this revolution is AcSys Biometrics Corp., a joint venture between AND Corporation, inventor and developer of Holographic/Quantum Neural Technology, HNeT, and NEXUS, a diversified holding company. NEXUS is a Burlington, Ontario-based company operating through a highly diversified web of autonomous subsidiaries and partnerships. (59) AcSys is a provider of one of the most advanced facial recognition systems on the market. AcSys’ Face Recognition System (FRS) is being incorporated into Nortel’s own product portfolio. AcSys’ FRS approach to security applies a proprietary technology for quick and reliable determination of human identity. It provides a scalable solution that integrates easily with existing systems and applications using standard network protocols.

In relation to Nortel’s Personal Internet strategy, Rick Collins, Senior Manager of Nortel’s "ProtoNet Project (Disruptive Solutions Implementation)," said of AcSys’ FRS:

"Layering AcSys’ face recognition’s capabilities within Nortel Networks’ solutions will make communication networks more personal. In the future, people may be recognized at a location, instead of logging in for some mobility services. I envision a network that knows who you are, and when you tell it, where you are, and can reach you whether you’re on your mobile phone or at your desktop." (60)

AcSys Biometrics face and speech recognition system is based on a patented core of artificial intelligence called Holographic/Quantum Neural Technology (HNeT).

HNeT neural networks could power a range of Nortel’s applications in the financial, manufacturing, security, surveillance, and medical sectors. With technologies like HNeT, with rates of learning up to 200 times faster than conventional neural nets, facial recognition via CCTV becomes a reality and countries with national CCTV infrastructures will view such technology as a natural extension of their networks. As with the example of the traffic control systems installed in Tiananmen Square and Lhasa, this process of extending a system’s surveillance capabilities is one of subtle erosion of rights. The dynamics of this process: continuously upgrading technology and incorporating functions unintended by the design.

« previous | 5 | 6 | 7 | 8 | 9 | 10 | 11 | 12 | 13 | 14 | 15 | 16 | 17 | 18 | next »