China's buildings are watching how people shop, cook and steal

Tongzhou Wanda Plaza in Beijing, one of the malls run by Dalian Wanda Group, which has installed cameras that use behaviour-recognition technology to track shoppers' movements in its malls, such as how long a person lingers in a store and whether he
Tongzhou Wanda Plaza in Beijing, one of the malls run by Dalian Wanda Group, which has installed cameras that use behaviour-recognition technology to track shoppers' movements in its malls, such as how long a person lingers in a store and whether he walks out with a bag in his hand. The technology captures shopping patterns, helping the landlord better optimise merchant layouts.PHOTO: BLOOMBERG

Proptech uses sensors to capture data to improve shopping and lifestyle experiences

BEIJING/SHANGHAI • Fintech is changing the way people borrow, invest and pay for things. But there is another type of technology - most noticeably in China - that is altering the way urban dwellers interact with their living and shopping environments.

It is property technology, or proptech, the use of new technologies like big data and machine learning to help individuals and companies buy, sell and manage real estate.

According to Jones Lang LaSalle, investment in proptech start-ups from 2013 to 2017 totalled US$7.8 billion, with China accounting for about 36 per cent. Last year, that figure came to almost US$20 billion (S$27 billion), data from market research firm Venture Scanner shows.

"In China, there is a very dynamic proptech ecosystem, quite mature and advanced at all levels," Jones Lang LaSalle's Asia chief executive officer Anthony Couse said at the firm's first proptech forum in Beijing in May. "Some say we're slow-moving, conservative," he said, referring to the real-estate industry. "I don't think that applies to us here in China."

One oft-cited reason for China's proptech leadership is that the country tends to put more emphasis on convenience than privacy.

That makes it easier for property companies to use transaction databases, facial-recognition cameras and other technology to improve people's shopping and living experiences, though developers still have to be mindful how they access and use personal data to avoid allegations of overreach.

Here is a snapshot of some of the ways that Chinese developers are using the Internet of Things (IoT), artificial intelligence and big data to improve people's living and shopping experiences.


Dalian Wanda Group upgraded its property management platforms at two Wanda Plazas by installing cameras that use behaviour-recognition technology to track shoppers' movements inside the mall, such as how long a person lingers in a store and whether he walks out with a bag in his hand.​

  • $10b Total investment in proptech start-ups from 2013 to 2017, with China accounting for about 36 per cent, according to Jones Lang LaSalle.

    10,000 Number of sensors in each of the 280 Wanda Plazas across China.

    62% The percentage by which labour costs on facilities maintenance was reduced in Longfor Group Holdings' residential communities after sensors were installed.

The technology, from Wanda's Huiyun management system, allows the group to capture and analyse a person's age, gender and shopping patterns, letting the landlord better optimise merchant layouts.

Shoppers do not know it, but they are also assigned a computerised ID so that they are recognised on their next visit.

Sensitive information, including people's facial images, is not stored to prevent the risk of personal privacy infringement issues and potential legal disputes, the company says.

The Huiyun technology now tracks conditions in about 280 Wanda Plazas across China, using more than 10,000 sensors in each of them. The sensors also detect heat and moisture levels, air quality and energy consumption. They are able to compare signals from a cook and the range hood in a restaurant, for example, and flash alerts when a stove is left unattended without being switched off.

Dalian Wanda also uses proptech when designing its shopping malls. The software can construct a 3D model using inputs from the designer and contractor throughout the construction process to maximise efficiency and ensure quality.

The technology enables the company to turn a bare site into a shopping mall, with every store open, within 18 months of breaking ground. That compares to around two years of construction alone for a project of a similar size.


Putting sensors around its residential communities has allowed Longfor Group Holdings to shift away from a labour-heavy property management system. Developed by the Beijing-based home builder's service arm using Internet-of-Things technology, the system has trimmed labour costs on facilities maintenance by as much as 62 per cent.

Some 480,000 facilities with sensors at developments nationwide capture data and send it to a central processing facility. The information can range from lower apartment floors being soaked by summer storms, to an elevator being out of order, to a manhole cover shifting and potentially endangering passers-by.

If enough data creates a warning, alerts are sent out to maintenance workers in the field who can pick up the job, similar to Uber drivers accepting a booking.

Where manual adjustment is not required, automation kicks in. At some developments, water sprays are automatically triggered if sensors deduce that a plot of lawn is too dry. In one residential community in Chongqing, trash cans automatically compress garbage, paring sanitation workers' emptying frequency to a fifth of usual levels.

What does Longfor do with the employees its technology has put out of work? The company uses them to provide next-level service for residents. An executive who travels a lot can ask these newly trained butlers to feed her fish while she is away, or elderly people can be walked outside in their wheelchairs every morning. Residents can even rate these stewards, as Longfor calls them, using an app.


Using facial recognition technology, Shui On Land started to use an app called Inno for access control in its Shanghai office buildings. A surprising finding emerged at one city-centre tower: 70 per cent of workers were female. Hence, Shui On renovated one of its shopping malls - Xintiandi Plaza - below this office area, tailoring an entire five floors just for women.

Apart from adding new stores from New York-based, Taiwanese-born fashion designer Jason Wu and Israeli cosmetics brand AHAVA, two of the upper floors have been remodelled to resemble a fancy showroom: In one corner, a chic kitchenette extends to a home decoration store, while a shop for outbound travel site sits next to a bookstore.

To further attract female shoppers, Shui On uses big data for distributing coupons and mall maps.

First-time visitors enter their mobile number on a fourth-floor screen to link to their WeChat accounts, giving Shui On access to their buying habits on Tencent Holdings, WeChat's parent company. Immediately, discount coupons are sent direct to a person's WeChat Pay wallet, making a cup of coffee cheaper or car parking free. The screen also offers brand suggestions based on previous shopping habits. The whole process is pared with facial recognition technology, so second-time visitors just need to stand still in front of the screen.

Shui On is also using proptech in a trial to assist tenants. Sensors, for example, can capture how many times a hesitant customer picks a book up and sets it down - useful data a store can use to reorganise displays and accelerate sales.


At almost 1,000 residential projects managed by Country Garden Holdings' affiliated service units across China, video surveillance cameras capture footage three times a minute, gathering real-time pictures ranging from what guards are doing to whether non-residents are intruding upon private property. The pictures are then sliced and diced into numerous data bits and sent to a cloud-based artificial intelligence platform that the service arm co-developed with Tencent last year.

This picture data from properties spanning 500 million sq m can help staff monitor multiple situations in real time - more than could be captured from a few cameras. The analysis can trigger automatic maintenance alerts, such as a full trash bin needing to be emptied.

Should a child be lost, the computer algorithms use behaviour-analysis technology to help guards locate him sooner. It has also allowed Country Garden's services arm to cut down the number of management staff working at its residential projects by a third, according to its chief information officer Yuan Hongkai.


A version of this article appeared in the print edition of The Straits Times on July 08, 2019, with the headline 'China's buildings are watching how people shop, cook and steal'. Print Edition | Subscribe