During the big data epidemic.
Editor’s note: This article is from the WeChat public account ” Tech planet “(ID: tech618) , author: Ma micro ice.
A sudden outbreak is unexpected, and the daily ups and downs of the outbreak data affect everyone’s heart.
Big data such as the number of infected persons, close contacts, activity trajectories, and time nodes have become key information for epidemic prevention and control. Behind the much-anticipated data, there is a group of people who, although they are not front-line anti-epidemic personnel such as doctors and nurses, went retrograde to the most severe areas after the outbreak.
“I didn’t expect that when I returned to Wuhan, I never left, and I worked for more than 40 days.” Xu Ke, a data engineer at Haizhi Science and Technology, said this is a lot of data that went retrograde to the epidemic area. One of the technical staff.
On January 20th at 8 pm, Xu Ke, who had just taken annual leave, arrived at his home in Ezhou, Hubei, and was notified by the company while he was eating dinner.
“The epidemic outbreak in Wuhan requires some epidemic analysis and support. We have limited staff in Wuhan and need to be on duty within 8 hours. Do you have any questions?” After putting down the phone, Xu Ke told his family that he received an emergency. The mission will return to Wuhan the next day.
The next morning, Xu Ke simply packed his clothes and drove back to Wuhan for emergency standby.
At that time, various cities had not sounded the alarm. Even in the streets of Wuhan, not many people wore masks. Most people, like Xu Ke, were not aware of what happened later.
Immediately afterwards, Wuhan closed the city, and epidemic prevention and control throughout the country began screening and isolation based on numbers. During the three months of the epidemic, each set of data and each string of codes has played an important role.
A big data battle against epidemics has opened the curtain. This is a practical review of the value of big data and the “data man.”
Destined retrogrades
“Unprecedented”, this is the word used by Xu Ke when it comes to feelings about the epidemic.
After the “War Epidemic” started, Ali,Many Internet companies such as Tencent and Baidu have begun to send emergency supplies to Wuhan. When receiving the call for an emergency recall, Xu Ke didn’t think too much, thinking that he would come back in a few days as usual to work overtime to catch up with the project. But I did not expect that I have been fighting so far.
On January 20, Gao Yongbo, the person in charge of Haizhi Hubei, received a request and needed to dispatch a team to form a technical team to participate in the data analysis of Wuhan Epidemic Prevention Command.
The epidemic prevention command set up in the epidemic is composed of government departments, hospitals, health and epidemic prevention agencies, and health and health committees. The epidemic prevention command will issue some data verification and data analysis requirements. After the data engineers cooperate with relevant departments to complete the analysis, they will provide the analysis results to the command for decision-making.
“Because I was very busy years ago, at the beginning we were going to take a holiday in advance this year and let everyone take a rest. I didn’t expect to encounter an unexpected epidemic situation.” President Yang Juan of Haizhi told Tech Planet, After receiving such calls from various places, we immediately started internal discussions. We were also very entangled. After all, there are still risks and we are worried about the emotions of employees, but everyone basically has nothing to say.
Gao Yongbo, who received the urgent task, asked everyone in the work group who can come to Wuhan as soon as possible. “Some have already returned home, some have turned around halfway, and others in Wuhan do not plan to go home. “It didn’t take long for six colleagues to say that they could return immediately, and Xu Ke was the second person he notified.
“After receiving the notification, all six of us arrived within one day,” Gao Yongbo told Tech Planet (WeChat ID: tech618).
At 11 am on January 21, Xu Ke and 6 colleagues gathered in Wuhan to stand by. “I don’t know what I’m going to do this time, I just know that it’s related to the epidemic.” Xu Ke, who had just arrived in Wuhan, had no time to eat lunch, so he immediately attended the work meeting.
With just one hour of work docking, Xu Ke, who is good at data analysis, realized the seriousness of the epidemic at once. He knew that the work to be started would be an unprecedented challenge.
Government departments at all levels have hundreds of types of data, and the total storage is extremely large, scattered across different departments. At the same time, medical epidemic prevention agencies at all levels also have a large number of manual forms of primary epidemic data. These complicated and complicated data should be quickly formed into a set of efficient data access, cleaning, and processing mechanisms, converted into accurate epidemic prevention information, and transmitted to the epidemic prevention headquarters. For Gao Yongbo, the biggest pressure is time and each Life represented behind the data. Later, someone asked Gao Yongbo’s feelings at the time, “I have never had such an experience. Every minute in my mind is the word of human life Guantian.” Gao Yongbo said.
Deng Hualiang, the implementation director of Haizhi Network, believes that:It is the most difficult and the most stressful work. “It will be relatively complicated for the first time. After we have access to these data and the processing rules have been established, new data will be updated at any time in the follow-up. This process is automatic. “.
“In a short period of time, we need to access the data of various departments to do data processing, cleaning, and association to form a data model. The data transfer mechanism has just been established, and the model is also explored. The changes are frequent, and the required Very urgent. That is the first stage of an outbreak, and there is a lot to do to explore. “Xu Ke said.
Xu Ke and his colleagues work day and night. Starting from January 21st, the first close contact information release was started. Basically, it was closed at 3 am or 4 am, or even at 4 or 5 am. It’s time to start work again. The front-line “war epidemic” is urgent. Xu Ke and colleagues are thinking about how to reduce time and form data model applications faster.
The first “data defense line”
Confirmed cases, suspected cases, and hot diagnoses. If these patient data can be obtained at the first time, it will be used for subsequent research and analysis. For example, to find these close contacts, notify them in time, and send the data in time. In the hands of front-line community workers, it is vital.
Since January 21st, Haizhi has assisted relevant departments in various places to process billions of data every day through Haizhi’s big data mining system and issue tens of thousands of “B-type people” (close contacts). The information provides support for the accurate investigation of community workers, which was basically updated once every half an hour.
The data model for epidemic prevention and control is very important, and colleagues on the front line have felt unprecedented pressure. Gao Yongbo said, “The usual work will give a time, such as one or two days. But the same amount of work now will be compressed and completed in one hour.”
At the beginning, they were arranged to rest in a nearby hotel after work, but with the outbreak of urban control, outsiders were not allowed to stay in hotels, and all hotels were vacated to aid medical teams. Xu Ke and his colleagues set up a camp bed directly next to the office, and rested when they were tired.
The camp bed by the data engineer at his desk
“We can only distinguish between day and night, nothing else,” Gao Yongbo said.
The first step of the epidemic prevention command is to quicklyThis data comes together. Deng Hualiang told Tech Planet (WeChat ID: tech618): “After converging, we found that the data standards are not uniform and the data quality is uneven. Then the next job is to quickly organize and clean these data.” It is undoubtedly a huge workload and a high-intensity thing.
As the number of confirmed diagnoses continues to increase, Wuhan will be closed at 10 am on the 23rd. The country is most concerned about how many people are out in Wuhan? Where did you go and by what means of transportation? The daily data is cleared daily, and the information of various types of objects in the daily data must be reported on the same day. These original first-hand materials can only be summarized by the Wuhan team at the first time. Compared with epidemic prevention in other cities, the workload of Wuhan is multiplied by hundreds, thousands, or even tens of thousands.
Deng Hualiang said, “Once we have mastered the information of the emigrants, we have used big data modeling methods to build hundreds of analytical models. There are cross-validation models for the authenticity of the data, there are confirmed cases models for landings, and there are emigrants to find out. Models, landing models with close personal contact, etc. Then push the analysis data of these models to frontline epidemic prevention personnel, and quickly go to the ground for verification. It can be said that a model is a battlefield, and each battlefield is about life and death. “ < / span>
The temporary team has a small staff. In the face of huge amounts of data processing information, various problems will inevitably occur. “Because it is a multi-department and multi-system gathering in one place and sending it from one place, compared to the difficulty of technical support, there may be more problems in the coordination mechanism of the entire analysis and operation. In addition to solving technical problems every day More often, we have to communicate with different departments, remind the data to be reported, ask for feedback, and ensure the smooth progress of the work. In addition to the operation of the mechanism, it is to continuously improve the algorithms and functions and reduce the processing time of the technology as much as possible. With each compression, there is more time to deal with more problems. “Xu Ke explained to Tech Planet.
Because of the traffic control in Wuhan, the technicians of Haizhi Beijing Headquarters cannot reach the Wuhan site, but in order to ensure timely assistance to the Wuhan team, the technical colleagues are online 24 hours a day to remotely troubleshoot and solve problems.
On January 30, the company closely coordinated with the reinforcement of three technical backbones stationed in Wuhan for rotation. On the same day, Haizhi opened an intranet big data analysis platform account for 22,000 first-line epidemic prevention and control personnel nationwide, and provided free support for the analysis of first-line epidemic prevention and control data. Subsequently, from many closed villages and towns in Nanchang, Nanjing, and Hubei, more than a dozen Haizhi engineers concentrated on the frontlines of epidemics in Wuhan, Xiaogan, Huanggang, and Ezhou in just a few days.
A big data epidemic prevention wall was established.
Fire Line reinforcements attack
For about one month, 9 engineers rotate the shaft, and finally the epidemic prevention and control data in WuhanThe model is gradually stable. “Next, we need to support other cities in Hubei. The local engineers are not enough. We can only transfer people from the headquarters.” Gao Yongbo said to Tech Planet.
Zhai Shidan, director of research and development, said, “I am on the phone with Yongbo every day. I am afraid and expecting. I am afraid that there is a temporary problem in the system, which delays the analysis. I am looking forward to the newly launched functions according to the requirements in the front. Good news to save people. Just like this, sometimes I ca n’t sleep well all night, thinking about where I can start and do better. “
On February 19th, 12 square cabin hospitals were fully opened in Wuhan. After the medical resources were alleviated, the Wuhan Epidemic Prevention and Control Campaign launched a total offensive and the data volume doubled. At the request of the Wuhan Epidemic Prevention and Control Headquarters, an additional data analyst was urgently dispatched to support the FireWire. Lian Ming, Zhang Shunmin, Song Yanchao, the three “Data Man” took the initiative to ask for help.
Lian Ming is the person in charge of the Northern District and has been working in Beijing all year round. At 8 pm on February 18, Lian Ming received a call from the company’s vice president after work. “Brothers here in Wuhan have been fighting for 30 consecutive days. In order to ensure the health of everyone, we have to replace them like medical staff. We need you to support the epidemic prevention and control in Wuhan now. Will you leave tomorrow morning? OK?” Go to the Lian Ming of the notification and reply directly to “No problem, clear your luggage immediately.”
Lian Ming then immediately informed his colleagues in the department, “Is there an emergency to go to Wuhan for support? Is there any problem? Let’s start tomorrow morning.” No one hesitated. At about 7 am on February 19, Lian Ming and his colleagues came to Beijing West Railway Station. The crowded West Railway Station used to be only a few passengers, and they were far apart.
At 7:26, the three boarded the G71 high-speed rail from Beijing to Shenzhen. After the city was closed in Wuhan, the G71 did not stop in Wuhan, but when it was heard that it was a support staff sent to Wuhan, the relevant Wuhan authorities communicated in advance and coordinated the Wuhan station to specifically approve the G71 to stop in Wuhan.
Lian Ming and colleagues at the high-speed rail station
Lian Ming and his colleagues are in the same compartment, with 70% of the passengers in the compartment. In areas outside Hubei on February 19, the epidemic prevention and control has entered the middle and late stages. Some companies in Beijing have resumed work, but Wuhan is still under strict control. At 13 o’clock in the afternoon, the high-speed rail arrived in Wuhan. Lian Ming and his colleague held a stamped passport,g.36krcdn.com/20200314/v2_2b02f28e4e57433490f7cee07db24f64_img_000 “data-img-size-val =” 706,338 “>
Smart big screen renderings made by Haizhi Company
As Sherman Stein explains, mathematicalization is much more than “the process of calculating a bunch of numbers.” “When in a sudden, super complicated and continuously dynamic process, if there is no big data, there is no way to support decision-making. At this time, big data becomes the decision itself.” Yang Juan told Tech Planet (WeChat ID: tech618 ) Said.
After the outbreak, Internet technology companies have made use of their big data technology capabilities to quickly invest in the epidemic.
On January 21st, Dingxiangyuan launched the “Real-time Epidemic Situation” information page; on the 22nd, WeChat launched the real-time epidemic search function; on the 23rd, Tencent Health and Baidu Map launched “Fresh Diagnosis Maps” at the same time, and launched Baidu Migration Data, Tencent’s highlights are on the “War Pneumonia” channel and “Focus on Pneumonia”, as well as epidemic maps and prevention manuals; the AI algorithm developed by the Dali Hospital shortens the genetic analysis of suspected cases from several hours to half an hour, greatly reducing the diagnosis time.
“War Plague” is nearing its end, but it is not over yet. Frontline combat positions still need to be upheld. In this battle of “human lives are in the sky”, big data engineers have undertaken a major and special mission to protect the health of many people.
For about one month, 9 engineers rotate the shaft, and finally the epidemic prevention and control data in WuhanThe model is gradually stable. “Next, we need to support other cities in Hubei. The local engineers are not enough. We can only transfer people from the headquarters.” Gao Yongbo said to Tech Planet.
Zhai Shidan, director of research and development, said, “I am on the phone with Yongbo every day. I am afraid and expecting. I am afraid that there is a temporary problem in the system, which delays the analysis. I am looking forward to the newly launched functions according to the requirements in the front. Good news to save people. Just like this, sometimes I ca n’t sleep well all night, thinking about where I can start and do better. “
On February 19th, 12 square cabin hospitals were fully opened in Wuhan. After the medical resources were alleviated, the Wuhan Epidemic Prevention and Control Campaign launched a total offensive and the data volume doubled. At the request of the Wuhan Epidemic Prevention and Control Headquarters, an additional data analyst was urgently dispatched to support the FireWire. Lian Ming, Zhang Shunmin, Song Yanchao, the three “Data Man” took the initiative to ask for help.
Lian Ming is the person in charge of the Northern District and has been working in Beijing all year round. At 8 pm on February 18, Lian Ming received a call from the company’s vice president after work. “Brothers here in Wuhan have been fighting for 30 consecutive days. In order to ensure the health of everyone, we have to replace them like medical staff. We need you to support the epidemic prevention and control in Wuhan now. Will you leave tomorrow morning? OK?” Go to the Lian Ming of the notification and reply directly to “No problem, clear your luggage immediately.”
Lian Ming then immediately informed his colleagues in the department, “Is there an emergency to go to Wuhan for support? Is there any problem? Let’s start tomorrow morning.” No one hesitated. At about 7 am on February 19, Lian Ming and his colleagues came to Beijing West Railway Station. The crowded West Railway Station used to be only a few passengers, and they were far apart.
At 7:26, the three boarded the G71 high-speed rail from Beijing to Shenzhen. After the city was closed in Wuhan, the G71 did not stop in Wuhan, but when it was heard that it was a support staff sent to Wuhan, the relevant Wuhan authorities communicated in advance and coordinated the Wuhan station to specifically approve the G71 to stop in Wuhan.
Lian Ming and colleagues at the high-speed rail station
Lian Ming and his colleagues are in the same compartment, with 70% of the passengers in the compartment. In areas outside Hubei on February 19, the epidemic prevention and control has entered the middle and late stages. Some companies in Beijing have resumed work, but Wuhan is still under strict control. At 13 o’clock in the afternoon, the high-speed rail arrived in Wuhan. Lian Ming and his colleague held a stamped passport,g.36krcdn.com/20200314/v2_2b02f28e4e57433490f7cee07db24f64_img_000 “data-img-size-val =” 706,338 “>
Smart big screen renderings made by Haizhi Company
As Sherman Stein explains, mathematicalization is much more than “the process of calculating a bunch of numbers.” “When in a sudden, super complicated and continuously dynamic process, if there is no big data, there is no way to support decision-making. At this time, big data becomes the decision itself.” Yang Juan told Tech Planet (WeChat ID: tech618 ) Said.
After the outbreak, Internet technology companies have made use of their big data technology capabilities to quickly invest in the epidemic.
On January 21st, Dingxiangyuan launched the “Real-time Epidemic Situation” information page; on the 22nd, WeChat launched the real-time epidemic search function; on the 23rd, Tencent Health and Baidu Map launched “Fresh Diagnosis Maps” at the same time, and launched Baidu Migration Data, Tencent’s highlights are on the “War Pneumonia” channel and “Focus on Pneumonia”, as well as epidemic maps and prevention manuals; the AI algorithm developed by the Dali Hospital shortens the genetic analysis of suspected cases from several hours to half an hour, greatly reducing the diagnosis time.
“War Plague” is nearing its end, but it is not over yet. Frontline combat positions still need to be upheld. In this battle of “human lives are in the sky”, big data engineers have undertaken a major and special mission to protect the health of many people.