Columbia University’s Data Science Institute, in collaboration with MediaTek, is developing AI architectures for smart cities that distribute tasks across local devices, edge servers, and cloud servers. Tested in West Harlem’s COSMOS testbed, this system optimizes urban functions like traffic management and emergency response. It also paves the way for 6G networks, enhancing AI applications and connectivity.
(Image Source: Columbia University, Data Science Institute)
Smart Cities Powered by AI and 6G Innovation: In the not-so-distant future, cities will pulse with intelligence powered by artificial intelligence (AI) integrated into their infrastructure. Smart sensors, cameras, and mobile devices will continuously analyze data to enhance safety, optimize traffic flow, and improve urban living. However, one significant challenge stands in the way of this vision: the processing power required to run AI models on devices like street cameras, sensors, and smartphones.
Columbia University’s Data Science Institute (DSI) is addressing this issue through the innovative work of its Sense, Collect, and Move Data Center. In collaboration with the semiconductor company MediaTek, researchers are developing cutting-edge AI architectures that intelligently distribute tasks across networks, ensuring efficiency and reliability.
Also Read: India tops charts with most mobile malware attacks
AI models require significant computational power, which traditional devices like cameras and sensors often lack. To solve this problem, DSI researchers, including Associate Professor Javad Ghaderi and Professor Gil Zussman from Columbia’s Electrical Engineering Department, are designing AI systems that allocate tasks across three layers of a network: local devices, edge servers, and powerful cloud servers.
“Our goal is to create a seamless system where AI tasks are automatically distributed based on current network conditions and the capabilities of each device,” explained Prof. Ghaderi. For example, a street camera could detect objects initially, while more complex analyses, like identifying suspicious activity, are sent to an edge server or cloud server for processing. This approach reduces latency, speeds up response times, and ensures critical data is processed efficiently.
To test these distributed AI architectures, the team is using the COSMOS testbed in West Harlem. COSMOS, part of the NSF Platforms for Advanced Wireless Research (PAWR) initiative, offers a dynamic environment for developing and experimenting with next-generation wireless systems.
Images captured by COSMOS nodes, equipped with sub-6GHz and mmWave radios, cameras, and other sensors, are processed in real-time using edge servers. The system ensures high-speed data processing even when network connections are weak.
“This testbed provides us with the flexibility to experiment in a real-world urban setting, enabling us to refine our system for practical applications,” said Mashid Ghasemi, an Electrical Engineering PhD student contributing to the project.
If successful, this distributed AI system could revolutionize urban life. Cities equipped with this technology could experience dramatic improvements in:
Such advancements would not only enhance the quality of life in cities but also make urban spaces more sustainable and efficient.
Beyond smart cities, this project is laying the groundwork for the development of 6G communication networks. The next step in wireless technology, 6G, is essential for supporting increasingly powerful AI applications.
Professor Ghaderi emphasized that current AI-driven applications like ChatGPT and AI-powered social media tools are heavily reliant on cloud-based services. Their performance depends on network connectivity and server capacity. However, the distributed AI architectures being developed can make these applications more reliable, even in low-bandwidth conditions or when server traffic is high.
“These new systems will enable faster, more dependable AI services, ensuring that users can access powerful tools even in challenging network environments,” he said.
As a global leader in semiconductor solutions, MediaTek plays a pivotal role in this collaboration. The company’s expertise in developing cutting-edge technology for smartphones, smart homes, and 5G aligns with the goals of the project.
Dr. Doru Calin, AVP and Head of the U.S. 6G Wireless Research Center at MediaTek, highlighted the importance of the partnership: “MediaTek powers technology that connects and enriches everyday life. Our collaboration with Columbia University researchers is a crucial step toward laying the foundation for 6G communication systems and pioneering future AI technologies.”
MediaTek’s involvement ensures that the AI architectures being developed will align with industry needs and standards, bridging the gap between academic research and practical implementation.
The vision of a smart city powered by AI is more than just a technological leap—it represents a shift toward more inclusive and sustainable urban environments. By optimizing resource use, reducing inefficiencies, and enhancing safety, these systems promise a better quality of life for all residents.
Moreover, the project supports broader goals of innovation, such as fostering new applications in education, healthcare, and public services. It also aligns with global efforts to create more environmentally friendly and energy-efficient cities.
The collaboration between Columbia University’s DSI and MediaTek is paving the way for the future of AI and wireless communication. With distributed AI architectures, the team is not only addressing today’s technological challenges but also preparing cities and networks for the demands of tomorrow.
Also Read: From gridlock to growth: Oxford University hails PM Modi's PRAGATI mission for digital governance
As AI continues to evolve, projects like this will be at the forefront of transforming urban landscapes, making them smarter, more efficient, and more connected than ever before. From managing traffic to laying the foundation for 6G, this partnership exemplifies the power of innovation and collaboration in shaping the cities of the future. Follow Education Post News for more global updates.
Loading ...
Copyright© educationpost.in 2024 All Rights Reserved.
Designed and Developed by @Pyndertech