Northeastern University has expanded its Institute for Intelligent Networked Systems to London, creating a new hub that aims to accelerate research in artificial intelligence and wireless communications while strengthening partnerships across Europe. The institute, formerly known as the Institute for the Wireless Internet of Things, has grown rapidly since it launched at Northeastern’s Boston campus in 2019. Having secured 68 patents for new inventions, built more than 30 industry partnerships and conducted advanced wellness research since its inception in 2019 at Northeastern University’s Boston campus, the institute has opened its first research office outside of the U.S. Director Tommaso Melodia described the London opening as a transatlantic expansion that will help the institute “create an interface between the world of connectivity and the world of intelligence.”
The London hub becomes the institute’s third location, joining its Boston headquarters and a satellite office in Burlington, Massachusetts, and will focus on making wireless communications faster, more energy efficient and more secure. The U.K. office is being led by professors Bipin Rajendran, an expert in intelligent computing systems, and Osvaldo Simeone, a specialist in information engineering, who have collaborated for a decade and see their skills as complementary. Melodia said the pair will broaden the institute’s expertise with work spanning machine learning, neuromorphic computing inspired by the human brain, and quantum computing. Simeone said the U.K. hub’s two missions would be “developing reliable artificial intelligence” and pursuing “emerging computing for our times,” and he is already heading a team of about 20 researchers based at Portsoken One on Northeastern’s London campus.
The London team is anchoring several major research efforts, including collaborations with U.S. technology giant IBM to develop the next generation of artificial intelligence hardware beyond silicon-based semiconductors. Simeone has acquired funding from the European Research Council to investigate the reliability of artificial intelligence in telecommunications, and he is hiring researchers for a project expected to start in February. He also has funding from the American not-for-profit organization Coefficient Giving to detect when large language models misbehave, using probes inside models such as those behind ChatGPT, Gemini and Copilot to identify when they attempt to cheat, jailbreak or deviate from intended behavior. The U.K.’s Advanced Research and Invention Agency has commissioned Rajendran, with Simeone also part of the three-year project, to look at building accelerators for large language models, and Rajendran said “the aim is to reduce the cost of LLM training or model development by a factor of one hundred or so compared to the [current] graphics processing units.” The institute has attracted $130 million in funding in seven years, has 225 members and works with more than 30 industry partners, and its London launch was marked by a conference in Devon House that brought together international experts and representatives from companies including IBM and Nvidia to signal that the new hub is open for collaboration.
