What developers can do on Earth Day 53

In the spring of March, everything is Su, and it is time to add greenery to the earth. However, with the progress of science and technology and social development, human consumption of the earth's resources, the impact on the ecological environment is also increasing. Based on this, this year, Microsoft and Intel have launched a green action for enterprise developers again - the 2022 Microsoft X Intel Hackathon Contest has started again!

---

The competition is held every year, how is this year different?

2020, the 51st Earth Day

Microsoft and Intel jointly launched the "AI Loves the Earth" series of technology sharing to help developers learn AI technology and empower sustainable development!

▲2020 Microsoft X Intel Online AI Love Earth Series Technology Sharing Summit

2021, the 52nd Earth Day "AI for Earth - Heroic Dreams Gather, Micro AI Combat" Hackathon Contest calls on all developers to protect the earth with technology through technical exchanges and advancement!

▲ Group photo of the 2021 Microsoft X Intel Hackathon Competition

2022, the 53rd Earth Day

Microsoft and Intel have teamed up again to launch a hackathon with the theme of cloud car racing and "work".

This year's competition focuses on the two topics of "Autonomous Driving" and "Industrial Production Safety". From March 10th to April 22nd, enterprise developers (enterprises as a unit) are sincerely invited to participate in the competition, using the OpenVINO™ tool suite, Intel ® Edge Insights for Industrial (EII) and Azure AI (AML) & Dv5 VM instance key technologies, develop more innovative green projects, help developers apply what they have learned, and jointly open the future of digital intelligence in the industry.

▌Autonomous driving: Using green technology to draw the outline of future travel

The data show that transport greenhouse gas emissions account for 16% of the total global emissions of 51 billion tons. As an important force for emission reduction, autonomous driving makes transportation and travel more efficient and greener with features such as intelligent route planning and zero carbon emissions.

Autonomous driving is a huge and complex project, and the following technologies and concepts cannot be avoided in research and development:

• Automotive Open System Architecture (AUTOSAR): The basic software modules provided by the AUTOSAR layered software architecture can be used in vehicles from different manufacturers and electronic components from different suppliers, which can help car companies reduce R&D expenditures.
• Advanced Driver Assistance System (ADAS): Effectively reduce driver traffic accidents through hazard prediction, traffic status update and driving behavior analysis.
• Software Defined Vehicles (SDV, Software Defined Vehicles): Under the intelligent air, the integration of in-vehicle software is more and more important to improve the experience of passengers and drivers. The traditional automotive electronic and electrical architecture ECU and ECU communicate through point-to-point, while the SOA software architecture divides the vehicle functions into different service components to solve the problem that the change of a certain functional component causes the whole body to be affected.
• Over-the-Air Technology (OTA, or Over-the-Air Technology): In the era of software-defined automobiles, by using the Internet of Things technology, the remote predictive maintenance, management and update of automobile software can be realized in the air interface of mobile communication. Help OEMs preset various models to enhance ADAS applications.
• Software-in-the-Loop Testing (SIL, Software-in-the-Loop) & Hardware-in-the-Loop Testing (HIL, Hardware-in-the-Loop): SIL can be deployed on a large number of servers without considering the target hardware, and the cost is relatively low Low. HIL test is different from SIL, which needs to consider the target hardware, the cost is high, and it is generally not deployed in large quantities. Industrialization of driving.

According to the degree of automation, relevant institutions at home and abroad divide autonomous driving into six levels from L0 to L5. The industry generally takes L3 as the watershed, the following is assisted driving, and the above is advanced autonomous driving. At present, the level of autonomous driving in the domestic market is mostly between L3 and L4. The most direct difference between L3 and L4 is that the latter does not require the intervention of the driver, and can achieve ultra-high autonomous driving in predictable environments such as high-speed and parks.

At present, the extremely complex traffic environment in my country is an important reason for restraining the development of autonomous driving. In terms of road condition recognition, vehicles need to complete the control and recognition of road obstacles, pedestrians, traffic signals, and other vehicle states. So far, it is possible to achieve this effect. Self-driving cars are rare. Microsoft and Intel hope that through this competition, companies with innovative strength will gather to use the OpenVINO™ tool suite and Microsoft Azure - AI (AML) technology to conduct autonomous driving model training, and show the training on a 1:16 physical frame As a result, it demonstrated the black technology of autonomous driving and helped to upgrade China's autonomous driving technology.

▌Industrial production safety: polish the "eyes of safety" in a low-carbon way

According to the statistics of Jiaziguangnian, industrial visual inspection is expected to exceed 50 billion yuan by 2026. Under the trend, more and more manufacturing companies are based on machine vision technology . By collecting and analyzing the data of personnel, equipment, production materials, environment, and processes in production, it can avoid improper operation of workers, visual fatigue of inspectors, and lead to inspection errors and other production safety risks, reduce costs, and improve inspection efficiency .

Due to the limitations of product differentiation, production environment and other conditions, different application scenarios have different requirements on the accuracy and delay of detection. Such important detection scenarios as chemical leakage detection and power facility detection often require 24-hour automatic monitoring, while sewage detection and industrial accumulation detection are usually not 24/7 detection.
With the advent of the cloud era, with the help of increasingly mature cloud technology, enterprises can build a cloud-edge collaborative industrial visual inspection platform, complete inference calculations in edge devices, perform unified management of cloud-to-edge terminals, reduce manual on-duty, and achieve accurate detection, reduce environmental pollution.

Schedule

This competition consists of two competition questions, "Autonomous Driving" and "Industrial Production Safety", covering five stages: competition registration, online training, submission of works, offline finals, and awards. The specific schedule is as follows:

1. Contest registration: March 10-April 22
2. Online training: April 25 - May 19
3. Online submission & preliminary review: May 6-May 9
4. Offline Finals: Time: May 20-May 22, Location: Shanghai
5. Result announcement & award ceremony: May 23-June 30
   *Due to the requirements of epidemic prevention and control in Shanghai and other cities, the time of the offline finals may be adjusted to some extent. The specific adjustment results will be notified separately. Please pay attention to the official website of the competition ( https://aka.ms/2022AIoThack ).

Challenge requirements

▌Autonomous driving

• Based on computer vision (classification and target detection), using a 1:16 frame, combined with the advanced architecture of Intel and Microsoft Azure, to conduct automatic driving competitions on designated tracks, including training, obstacle avoidance, acceleration/deceleration, cornering, Stop Sign etc. scene.
• Key enterprises: OEMs, OEMs, Microsoft/Intel global partners and related solution providers in the automotive industry
• The participating teams must use the hardware platform customized by the organizing committee for development reference: Donkey Car
• Participating teams must possess the following technical abilities:
  Familiar with Python2.0 / 3.0
  Familiar with Linux- Ubuntu 18x/20x
  Familiar with VScode, WSL, Bash, etc.
  [Optional]Familiar with Arduino
  [Optional]Experience in OpenVINO™ Toolkit
  [Optional]Experience in Azure AI (AML)
  [Optional]Experience on TensorFlow 2.x/YOLO/Keras/etc.
• The participating teams must apply and deploy the following Intel and Microsoft Azure related products in their own solutions:
  OpenVINO™ Toolkit
  Microsoft Azure - AI (AML)

▌Industrial production safety

• Based on machine vision technology, through real-time AI detection of participants, equipment or materials in the production environment, it can improve production safety in the industrial field and reduce environmental pollution.
• Teams must use the hardware platform customized by the organizing committee for development, deployment and demonstration:
  Industrial MV Development Kit
• Teams are recommended to possess the following technical abilities:
  Familiar with Python/C++
  Familiar with Linux/Docker
  Experience in AI Inference Deployment
  Experience in Azure AI/IOT service
• The participating teams must apply and deploy the following Intel and Microsoft Azure related products in their own solutions:

OpenVINO™ Toolkit
Intel® Edge Insights for Industrial (EII)
Microsoft Azure – Azure IoT
Microsoft Azure Dv5 VM – Series instances based on ICX

Enterprises show cool skills, and the harvest is more promising

In-depth participation in technical competition questions, hacking with experts, and demonstrating technical strength... In fact, for participating companies, the benefits of participating in the event are far more than that! Participate in the competition and become a green ambassador company to get:

▌Product & Service Publicity Exposure

• Participating companies will enjoy free online and offline multi-channel traffic exposure and promotion of this competition to enhance the awareness of brands and core business technologies.
• The winning team in the competition will participate in the live awards during the Microsoft Build 2022 conference and demonstrate their works/core technical capabilities.

▌Communicate with peer customers and big coffee

• In-depth exchange of technology with customers and partners in the same industry, sharing corporate experience and product characteristics, getting to know more industry and technical friends, and expanding the technology circle and business market.
• Face-to-face with industry experts and technical judges, in-depth exchanges on the industry frontier trends of artificial intelligence and the Internet of Things, and help enterprises to innovate in technology.
  ▌Original technical support to accelerate technological innovation
• Microsoft global AIoT SWAT experts and Intel IoT technical experts guide the whole process, participate in and provide original technical support, and help enterprise development teams to quickly deploy actual combat.
• Free use of Microsoft Azure and OpenVINO™ toolkit throughout the entire process, and deploy and test models on the competition's customized autonomous driving/industrial vision development kit.
  The best way to see the future is to create it. Scan the code to follow the official website of the competition, learn more about the dynamic information of the competition, add greenery to the earth, and promote the innovation and upgrading of the industry.

Browse more contest information

---

Scan the code to follow Microsoft China MSDN to get more cutting-edge technology information and learning content!