Alibaba Cloud Announces Machine Learning platform PAI 2.0, Driving Innovations in Life Sciences and Manufacturing

March 29, 2017

Alibaba Cloud, the cloud computing arm of Alibaba Group, announced today that it has upgraded its machine learning platform PAI, at its Computing Conference Shenzhen Summit. Built on large-scale, high-performance and distributed computing, PAI 2.0 will help customers easily deploy large-scale data mining and modeling. Its advanced and diversified artificial intelligence (AI) algorithms are aimed at empowering innovations in the life science and manufacturing sectors.

"In the past year, Alibaba Cloud has implemented a number of real-life AI applications for customers across industries. As China's first machine learning platform, PAI is born to make our AI program, an effective tool for helping people and businesses to resolve practical issues. We are also seeing our data intelligence capability being leveraged to assist customers in creating better healthcare services and endorsing smart manufacturing in China," said Dr. Jingren Zhou, Chief Scientist of Alibaba Cloud.

Alibaba Cloud upgrades machine learning platform PAI

Alibaba Cloud's PAI 2.0 significantly reduces development costs and entry barriers of AI for users.

Key features of PAI 2.0 include:
Diversified and innovative algorithms: PAI has more than 100 sets of algorithms covering data preprocessing, neural network, regression, classification, predictions, evaluations, statistical analysis, feature engineering and deep-learning architecture.
Deep-learning architecture: The entire computing architecture is optimized for different deep learning frameworks. It also supports a one-click function to deploy application program interface (API), solving the challenge of modeling and service integration.
Large-scale computing power: Powered by Apsara, Alibaba Cloud's large-scale computing engine, PAI provides ultra-large-scale distributed computing capabilities that allow organizations to handle petabyte-level computing tasks daily.
User-friendly interface: PAI's data visualization capabilities allow developers to quickly include components into workflows by easy drag and drop features, greatly enhancing the efficiency of model building and debugging.

Through the platform's flexible support, organizations can easily build solutions for a variety of application scenarios, such as tumor diagnosis and data-driven production systems.

ET Medical Brain: Delivering the AI treatment to healthcare challenges

On top of releasing PAI 2.0, Alibaba Cloud has also developed ET Medical Brain, a suite of AI solutions built on its cloud computing infrastructure to drive advancements in healthcare sector. With processing a wealth of medical knowledge and experience, ET Medical Brain aims to ease the workload of medical personnel by taking roles of patients' virtual assistants, and participating in medical imaging, drug development, hospital management and wearable healthcare devices.

Medicine as a knowledge-driven discipline requires huge quantities of data to be processed from different sources. AI allows computers to play a significant role in assisting decision-making in the context of healthcare, freeing medical personnel from complex repetitive tasks to focus on delivering quality care to patients.

In addition to Alibaba Cloud's own development in AI healthcare solutions, the company is also bringing together medical and IT experts to collectively improve China's population health. Alibaba Cloud, Intel, and LinkDoc (an oncology big data company) will co-host the 2017 Tianchi AI competition for healthcare promote real-world study and AI supported diagnosis in medical field. Participants will compete by developing algorithms and models to analyze high-resolution chest CT scan data sets and screen the pulmonary nodules through data intelligence. The competition aims to improve the accuracy of early lung cancer detection and stimulate the integration of traditional treatments and machine learning. Registration for the competition will open during the Shenzhen Summit on 29 March 2017.

ET Industrial Brain: Making smart manufacturing smarter

Alibaba Cloud is also empowering enterprises with ET Industrial Brain, the company's suite of AI solutions in the field of industrial production. In line with the national initiative "Made in China 2025", ET Industrial Brain aims to help companies to centralize data management of manufacturing process with Alibaba Cloud's expertise in computing and data analytics:

Malfunction prediction: ET Industrial Brain helps manufacturers to plan ahead for maintenance in order to prevent large-scale mechanical failures and uncertainty caused by the loss from downtime and to improve equipment utilization and the efficiency of spare parts stocks.
Real-time monitoring: ET Industrial Brain monitors the entire 360 degree production process, which can enhance overall product yields with timely interventions.
Energy saving: Data analytics applied to production data can help to optimize energy consumption, reducing costs and carbon emissions.
Craftsmanship suggestion: AI solutions make recommendations to workers to fix vulnerabilities in the manufacturing process, in order to allow companies to be more productive.

With these new cutting-edge developments, Alibaba Cloud is a leader in harnessing AI and deep learning technologies to expand the innovative possibilities across industries and enable companies to be more productive and cost-effective.