JD Cool Stuff #2: Trash Sorting Apps and How Smart Cities Are Becoming Operating Systems (2 of 3)


This is Part 2 on my list of cool stuff JD is doing. Not necessarily the big initiatives – but the smaller and often more inventive ideas. In Part 1, I talked about their farming initiative, which is pretty intriguing.

In this part, I look at their trash sorting program and how China’s smart cities are starting to look like operating systems (i.e., digital platforms).


China Is the World’s Laboratory for Smart Cities

I’ve been arguing that the next big thing out of digital China is going to be smart cities. And for lots of reasons.

  • There are lots and lots of pilot projects. About 500 of the world’s 1,000 smart city pilot projects are in China.
  • There are four particularly effective companies leading the charge. PingAn, Alibaba, Tencent and Huawei (i.e., PATH) are kind of the designated national champions for smart cities.
  • This is about combining digital and infrastructure. And China is really good at both of these. China builds the equivalent of two Chicago’s every year. Here is a cool list of some of the big infrastructure projects (30 giant Chinese infrastructure projects that are reshaping the world).
  • Chinese urbanization is continuing. A population the size of Japan moves into Chinese cities about every 8 years. And a lot of the cool digital China stuff, like on-demand delivery, depend on lots of people and high urban density.
  • Everyone is on a smartphone – and they are pretty enthusiastic adopters of digital stuff. Many smart city applications succeed only if they are widely adopted by the population.
  • There is a ferocious entrepreneurial environment, especially in mobile apps.
  • There is already tons of data on the consumer side.
  • The cost of cloud storage and processing are falling.

Hence, China is pretty much the world’s laboratory for smart cities.

Smart Cities Are Starting to Look Like Operating Systems (i.e., Digital Platforms).

Layer 1 of this is infrastructure, which is what everyone talks about. This is putting sensors in streets, pollution sensors in the water system, drones in the sky and cameras everywhere. Smart infrastructure (especially IoT) is happening in both the private and public sectors – and these gather information and are increasingly active. For example, smart street lights can react in real time. Lanes on the streets can reverse direction. Power can be re-routed in real time to different parts of the city. And so on.

But this is just the first version. which is putting various digital tools into the existing city infrastructure. Most roads and urban centers developed over decades, if not centuries. So they have lots of built in assumptions about transportation (cars, horses previously) and how people work and live. The next version of smart infrastructure could be building entirely new cities that operate differently.

Layer 2 of this is data, both public and private. Once you get everything digitized and connected, you start to gather lots of data. And you want to standardize the data and make it public. Some of this data will come from cameras and sensors. But it can also come from businesses and a billion consumers walking around with smartphones.

Layer 3 is software, apps and various digital tools that run on the platform. This is really the most powerful part. Once the platform is in place and the data (and APIs) are in place, millions of developers can start building apps that run on the platform. That’s when the smart city becomes an operating platform like your smartphone and PC. And this should really take off in China, given the highly entrepreneurial environment.

The early action is in security, transportation, waste and water. But there can also be apps for community, health, government services and other areas. And this is where the discussion starts to be cultural and political. Depending on what country I’m in, I hear different concerns about the security / surveillance aspect of smart cities. In Germany and Poland, the surveillance concern is brought up immediately. If I’m in Mexico or Brazil (where crime is rampant), the possibility of decreasing crime is always discussed.

Example: Alibaba Launches a City Brain in Hangzhou

Alibaba launched its City Brain project in Hangzhou in 2016 with a big focus on transportation. Street lights became smart. Traffic cameras began watching everything. And everything was connected to the cloud. It uses data from video footage, traffic bureaus, and public transportation systems and mapping apps. According to the press releases, their system can make live traffic predictions, optimize traffic flow, and detect traffic incidents. As a result, Hangzhou dropped from 5th to 57th on the list of China’s most congested cities.

Hangzhou City Brain 2.0. was launched in 2018, now covering 42 square kilometers and with the ability to report traffic violations with 95% accuracy (per their report). It has 1,300 traffic lights controlled by AI and +200 police available through the platform to attend to traffic emergencies. In theory, City Brain should be able to learn traffic patterns and make recommendations to improve traffic efficiency, like the best ways to plan new roads. Or changes to bus routes.

City Brain is now being expanded to Suzhou, Guangzhou, Macau and Malaysia. 

JD Launches a Cool Trash Sorting App

Which brings me back to the cool JD project.

Earlier this year, JD announced a new trash sorting app for consumers in parts of Beijing. It was a pilot-type program in response to some pretty serious regulations happening in China about trash. Basically, the landfills of China are filling up way ahead of schedule and the government is getting pretty active.

The new regulations, started in Beijing and Shanghai, make trash sorting mandatory. They are expected to be expanded across China. Specifically, it requires all trash to be sorted into wet waste (i.e., stuff from the kitchen), dry waste, recyclables and hazardous waste.

So how would a smart city tackle this question? As opposed to a more traditional city?

I was visiting the Huawei campus a few months ago and got a good look at a lot of their projects. And I chatted with the head of their smart cities initiatives. Unsurprisingly, a lot of their focused on security and transportation. Transportation, in particular, is a field that is well suited to 5G connectivity.

But one of their smaller projects was putting IoT sensors in garbage cans. Which lets the system know how full each can is. And if you know how full each garbage can is, you can then re-route the garbage trucks based on real-time need. That’s pretty interesting. So much of the cool smart city stuff is really about going from fixed to more dynamic activities. Instead of the buses all taking the same routes, they can change based on traffic and demand. Instead of police all walking the same routes, you can re-direct them. And instead of garbage trucks taking the same routes, you set them dynamically.

JD’s idea for this was to give consumers a mobile app that is a trash sorting utility – and that can also “gamify” the activity. From their press releases, here is what I think they are doing in about 19 communities in the south of Beijing this year.

  • Consumers buy small green bins with QR codes. These are placed outside their apartments. And, of course, they can be bought on JD.com
  • Each bin can be linked to a WeChat account of a person in the family. So each family is putting their waste in their own personal bin. For recyclables, families are given a bag with a QR code, which is also linked to the WeChat account.
  • The JD mobile app lets users identify garbage by type (wet, dry, recyclables, hazardous) for proper sorting. They just take a picture of the item and it tells them where it goes. Computer vision is something that AI does quite well and the more people taking pictures, the smarter the system should get at identifying types of garbage. I have also heard there is a natural language processing version of this that can be done by voice. But I don’t know the status of that (it sounds iffy).
  • So consumers should, in theory, should put all the trash into the right bins. And then the trash and recycling companies come by and do pick-up. And they can check who is doing it correctly, which earns the families points.

It’s kind of intriguing. It’s a smart city project but it is using a mix of consumer engagement and city resources (garbage trucks). For consumers, a mix of a useful utility (trash identification) and gamification (earning points). I like services like this. The useful or needed utility (WeChat messenger, Mobike, most Google products) gets the users into the system. And the gamification aspects keep them there.

Final Question: Are Cities Going to Become a New Type of Digital Platform?

In my Asia Tech Class, I had an introductory lecture on digital platforms (and network effects) and why they are so important. Some argue there are only two types of digital platforms: innovation platforms and marketplace platforms.

  • Innovation platforms are where developers or content creators use the platform to create something new (like developers creating apps for the app store or YouTube creators making videos). So think PCs and smartphone operating systems.
  • Marketplace platforms are for transactions (which can be monetary, social and other) so that’s Didi / Uber in mobility services, Taobao / JD in consumer products and Meituan / Grab in consumer services.

I think there are more like 6 major platform types. I think payments are a unique type. As are audience builders, like YouTube. There are standardization and coordination platforms (like Slack). And it also gets more complicated as you go beyond having just two user groups. Or when you have complementary platforms. But I’ll go into that stuff in another talk.

My question here is: are smart cities a new type of digital platform?

  • Definitely app developers are going to be involved. So that would make it like an innovation platform in its earliest stages. But as the apps move from security and transportation to culture, education and community it could be very different.
  • These are probably local platforms, not the global platforms (AWS, Expedia, YouTube) everyone gets excited about. And not the regional giants of China (Alibaba, Tencent). These would be pretty local.
  • And they would be digital-physical hybrids (discussed in podcast about Vipshop), which means they would not be as scalable. But would probably be more defendable.

It’s an interesting question. We’ll see what happens.


Anyways, that’s it for Part 2. Next, I’m going to look at JD’s logistics and CPG projects.

Cheers and thanks for subscribing,



I write, speak and consult about how to win (and not lose) in digital strategy and transformation.

I am the founder of TechMoat Consulting, a boutique consulting firm that helps retailers, brands, and technology companies exploit digital change to grow faster, innovate better and build digital moats. Get in touch here.

My book series Moats and Marathons is one-of-a-kind framework for building and measuring competitive advantages in digital businesses.

Note: This content (articles, podcasts, website info) is not investment advice. The information and opinions from me and any guests may be incorrect. The numbers and information may be wrong. The views expressed may no longer be relevant or accurate. Investing is risky. Do your own research.


Leave a Reply