̽��ֱ�� of Cambridge - Nic Lane

Building business partnerships in AI, quantum, cybersecurity and computer architecture

skbf2 — Wed, 18 Sep 2024 14:25:55 +0000

Hear from four of our leading researchers on their work and why partnering with industry is key to their success.

Three Cambridge researchers awarded Royal Academy of Engineering Chair in Emerging Technologies

sc604 — Thu, 14 Mar 2024 11:06:52 +0000

From atomically thin semiconductors for more energy-efficient electronics, to harnessing the power of the sun by upcycling biomass and plastic waste into sustainable chemicals, their research encompasses a variety of technological advances with the potential to deliver wide-ranging benefits.

Funded by the UK Department for Science, Innovation and Technology, the Academy’s Chair in Emerging Technologies scheme aims to identify global research visionaries and provide them with long-term support. Each £2,500,000 award covers employment and research costs, enabling each researcher to focus on advancing their technology to application in a strategic manner for up to 10 years.

Since 2017, the Chair in Emerging Technologies programme has awarded over £100 million to Chairs in 16 universities located across the UK. Of the four Chairs awarded in this round, three were awarded to Cambridge researchers.

Professor Manish Chhowalla FREng, from the Department of Materials Science and Metallurgy, is developing ultra-low-power electronics based on wafer-scale manufacture of atomically thin (or 2D) semiconductors. ̽��ֱ��atomically thin nature of the 2D semiconductors makes them ideal for energy-efficient electronics. To reap their benefits, complementary metal oxide semiconductor processes will be developed for integration into ultra-low power devices.

Professor Nic Lane and his team at the Department of Computer Science and Technology, are working to make the development of AI more democratic by focusing on AI methods that are less centralised and more collaborative, and offer better privacy protection.

Their project, nicknamed DANTE, aims to encourage wider and more active participation across society in the development and adoption of AI techniques.

“Artificial intelligence (AI) is evolving towards a situation where only a handful of the largest companies in the world can participate,” said Lane. “Given the importance of this technology to society this trajectory must be changed. We aim to invent, popularise and commercialise core new scientific breakthroughs that will enable AI technology in the future to be far more collaborative, distributed and open than it is today.”

̽��ֱ��project will focus on developing decentralised forms of AI that facilitate the collaborative study, invention, development and deployment of machine learning products and methods, primarily between collections of companies and organisations. An underlying mission of DANTE is to facilitate advanced AI technology remaining available for adoption in the public sphere, for example in hospitals, public policy, and energy and transit infrastructure.

Professor Erwin Reisner, from the Yusuf Hamied Department of Chemistry, is developing a technology, called solar reforming, that creates sustainable fuels and chemicals from biomass and plastic waste. This solar-powered technology uses only waste, water and air as ingredients, and the sun powers a catalyst to produce green hydrogen fuel and platform chemicals to decarbonise the transport and chemical sectors. A recent review in Nature Reviews Chemistry gives an overview of plans for the technology.

“ ̽��ֱ��generous long-term support provided by the Royal Academy of Engineering will be the critical driver for our ambitions to engineer, scale and ultimately commercialise our solar chemical technology,” said Reisner. “ ̽��ֱ��timing for this support is perfect, as my team has recently demonstrated several prototypes for upcycling biomass and plastic waste using sunlight, and we have excellent momentum to grasp the opportunities arising from developing these new technologies. I also hope to use this Chair to leverage further support to establish a circular chemistry centre in Cambridge to tackle our biggest sustainability challenges.”

“I am excited to announce this latest round of Chairs in Emerging Technology,” said Dr Andrew Clark, Executive Director, Programmes, at the Royal Academy of Engineering. “ ̽��ֱ��mid-term reviews of the previous rounds of Chairs are providing encouraging evidence that long-term funding of this nature helps to bring the groundbreaking and influential ideas of visionary engineers to fruition. I look forward to seeing the impacts of these four exceptionally talented individuals.”

Three Cambridge researchers – Professors Manish Chhowalla, Nic Lane and Erwin Reisner – have each been awarded a Royal Academy of Engineering Chair in Emerging Technologies, to develop emerging technologies with high potential to deliver economic and social benefits to the UK.

̽��ֱ�� of Cambridge

L-R: Manish Chhowalla, Nic Lane, Erwin Reisner

̽��ֱ��text in this work is licensed under a Creative Commons Attribution-NonCommercial-ShareAlike 4.0 International License. Images, including our videos, are Copyright © ̽��ֱ�� of Cambridge and licensors/contributors as identified. All rights reserved. We make our image and video content available in a number of ways – on our main website under its Terms and conditions, and on a range of channels including social media that permit your use and sharing of our content under their respective Terms.

Yes

UK-US Summit for Democracy announces Cambridge team as joint winners of challenge to detect financial crime

cjb250 — Fri, 31 Mar 2023 12:53:46 +0000

̽��ֱ��announcement came at the second UK-US Summit for Democracy on 30 March 2023. ̽��ֱ��prize challenges innovators on both sides of the Atlantic to build solutions that enable collaborative development of artificial intelligence (AI) models, while keeping sensitive information private.

Driven by a shared priority to employ data to help solve critical global challenges in a manner that supports US and UK commitments to democratic values and the fundamental right to privacy, the challenges focused on developing PETs solutions for two scenarios: forecasting pandemic infection and detecting financial crime.

A team led by Professor Nic Lane from the Department of Computer Science and Technology at the ̽��ֱ�� of Cambridge was named joint winner in the financial crime category. Their challenge was to develop a privacy-preserving solution to help tackle the challenge of international money laundering.

Xinchi Qiu, a PhD student in Professor Lane’s lab, said: “We developed an end-to-end privacy-preserving federated learning solution to detect potentially anomalous payments, leveraging a combination of inputs from a number of financial institution and different banks. Our project aims to develop a method that can utilise all the inputs from different institutions while protecting the original data.”

Professor Lane said: "Right now, machine learning with federated and other privacy preserving methods are niche. But in the near future they will be the norm. Most of the world's data is inaccessible for machine learning – however these new methods are making such data available in safe manner. This will be a game changer for many high impact domains that are currently starved of sufficient data, such as health, finance and legal. Our solution shows how this can be done effectively for money laundering, but our methods can migrate to these other domains."

Experts from academic institutions, global technology companies, and privacy start-ups competed for cash prizes from a combined UK-US prize pool of $1.6 million (£1.3 million). ̽��ֱ��winning solutions combined different PETs to allow the AI models to learn to make better predictions without exposing any sensitive data. This focus on combining privacy approaches encouraged the development of innovative solutions that address practical data privacy concerns in real world scenarios.

In the final phase of the challenges, the privacy guarantees of the solutions were put to the test by ‘red teams’, who attempted to reveal the original data used for training the models. ̽��ֱ��resilience of the solutions to these attacks determined the final winners.

Michelle Donelan, Secretary of State for the UK Department for Science, Innovation and Technology, said: “Never before has our privacy been so important and we must protect our democratic values by safeguarding the right to privacy. That is why the UK and its allies are collaborating to create innovative technologies that enable public institutions to combat financial crime and promote public health without compromising the confidentiality of the sensitive data they manage.”

UK participants also received support from the UK Information Commissioner’s Office to help them consider how their solutions could demonstrate compliance with key UK data protection regulation principles.

John Edwards, UK Information Commissioner, said: “Privacy enhancing technologies can help analyse data responsibly, lawfully and securely and it will be important for regulators and industry to continue to work together to support responsible innovation in these technologies.”

Arati Prabhakar, Assistant to the President for Science and Technology and Director of the White House Office of Science and Technology Policy, added: “Data has the power to drive solutions to some of our biggest shared challenges, but much of that data is sensitive and needs to be protected.”

Adapted from a press release from the Centre for Data Ethics and Innovation

A Cambridge team has been announced as one of the winners of a prize to drive ‘innovation in privacy-enhancing technologies that reinforce democratic values’ for its work on tackling international money laundering.

Most of the world's data is inaccessible for machine learning – however, these new methods are making such data available in a safe manner. This will be a game changer for many high impact domains

Nic Lane

Geralt

Networks

̽��ֱ��text in this work is licensed under a Creative Commons Attribution-NonCommercial-ShareAlike 4.0 International License. Images, including our videos, are Copyright © ̽��ֱ�� of Cambridge and licensors/contributors as identified. All rights reserved. We make our image and video content available in a number of ways – as here, on our main website under its Terms and conditions, and on a range of channels including social media that permit your use and sharing of our content under their respective Terms.

Yes

Licence type:

Public Domain

Can federated learning save the world?

sc604 — Mon, 10 May 2021 11:23:31 +0000

Artificial intelligence models are used increasingly widely in today’s world. Many carry out natural language processing tasks – such as language translation, predictive text and email spam filters. They are also used to empower smart assistants such as Siri and Alexa to ‘talk’ to us, and to operate driverless cars.

But to function well these models have to be trained on large sets of data, a process that includes carrying out many mathematical operations for every piece of data they are fed. And the data sets they are being trained on are getting ever larger: one recent natural language processing model was trained on a data set of 40 billion words.

As a result, the energy consumed by the training process is soaring. Most AI models are trained on specialised hardware in large data centres. According to a recent paper in the journal Science, the total amount of energy consumed by data centres made up about 1% of global energy use over the past decade – equalling roughly 18 million US homes. And in 2019, a group of researchers at the ̽��ֱ�� of Massachusetts estimated that training one large AI model used in natural language processing could generate around the same amount of CO₂ emissions as five cars would generate over their total lifetime.

Concerned by this, researchers in Cambridge's Department of Computer Science and Technology set out to investigate more energy-efficient approaches to training AI models. Working with collaborators at the ̽��ֱ�� of Oxford, ̽��ֱ�� College London, and Avignon Université, they explored the environmental impact of a different form of training – called federated learning – and discovered that it had a significantly greener impact.

Instead of training the models in data centres, federated learning involves training models across a large number of individual machines. ̽��ֱ��researchers found that this can lead to lower carbon emissions than traditional learning.

Senior Lecturer Dr Nic Lane explains how it works when the training is performed not inside large data centres but over thousands of mobile devices – such as smartphones – where the data is usually collected by the phone users themselves.

“An example of an application currently using federated learning is the next-word prediction in mobile phones,” he said. “Each smartphone trains a local model to predict which word the user will type next, based on their previous text messages. Once trained, these local models are then sent to a server. There, they are aggregated into a final model that will then be sent back to all users.”

And this method has important privacy benefits as well as environmental benefits, points out Dr Pedro Porto Buarque De Gusmao, a postdoctoral researcher working with Lane.

"Users might not want to share the content of their texts with a third party,” he said. “In federated learning, we can keep data local and use the collective power of millions of mobile devices together to train AI models without users’ raw data ever leaving the phone.”

“And besides these privacy-related gains,” said Lane, “in our recent research, we have shown that federated learning can also have a positive impact in reducing carbon emissions.

“Although smartphones have much less processing power than the hardware accelerators used in data centres, they don’t require as much cooling power as the accelerators do. That’s the benefit of distributing the training of models across a wide pool of devices.”

̽��ֱ��researchers recently co-authored a paper on this called ‘Can Federated Learning save the planet?’ and will be discussing their findings at an international research conference, the Flower Summit 2021, on 11 May.

In their paper, they offer the first-ever systematic study of the carbon footprint of federated learning. They measured the carbon footprint of a federated learning setup by training two models — one in image classification, the other in speech recognition – using a server and two chipsets popular in the simple devices that targeted by federated methods. They recorded the energy consumption during training, and how it might vary depending on where in the world the chipsets and server were located.

They found that while there was a difference between CO₂ emission factors among countries, federated learning under many common application settings was reliably ‘cleaner’ than centralised training.

Training a model to classify images in a large image dataset, they found any federated learning setup in France emitted less CO₂than any centralised setup in both China and the US. And in training the speech recognition model, federated learning was more efficient than centralised training in any country.

Such results are further supported by an expanded set of experiments in a follow-up study (‘A first look into the carbon footprint of federated learning’) by the same lab that explores an even wider variety of data sets and AI models. And this research also provides the beginnings of necessary formalism and algorithmic foundation of even lower carbon emissions for federated learning in the future.

Based on their research, the researchers have made available a first-of-its-kind ‘Federated Learning Carbon Calculator’ so that the public and other researchers can estimate how much CO₂is produced by any given pool of devices. It allows users to detail the number and type of devices they are using, which country they are in, which datasets and upload/download speeds they are using and the number of times each device will train on its own data before sending its model for aggregation.

They also offer a similar calculator for estimating the carbon emissions of centralised machine learning.

“ ̽��ֱ��development and usage of AI is playing an increasing role in the tragedy that is climate change,” said Lane, “and this problem will only worsen as this technology continues to proliferate through society. We urgently need to address this which is why we are keen to share our findings showing that federated learning methods can produce less CO₂ than data centres under important application scenarios.

“But even more importantly, our research also shines a light as to how federated learning should evolve towards being even more broadly environmentally friendly. Decentralized methods like this will be key in the invention of future sustainable forms of AI in the years ahead.”

̽��ֱ��researchers will be presenting their work at the Flower Summit 2021, which takes place on Tuesday 11 May 2021.
Can Federated Learning Save the Planet, Xinchi Qiu, Titouan Parcollet, Daniel J. Beutel, Taner Topal, Akhil Mathur, Nicholas D. Lane.
Federated Learning Carbon Calculator
A first look into the carbon footprint of federated learning, Xinchi Qiu, Titouan Parcollet, Javier Fernandez-Marques, Pedro Porto Buarque de Gusmao, Daniel J. Beutel, Taner Topal, Akhil Mathur, Nicholas D. Lane

Training the artificial intelligence models that underpin web search engines, power smart assistants and enable driverless cars, consumes megawatts of energy and generates worrying carbon dioxide emissions. But new ways of training these models are proven to be greener.

̽��ֱ��development and usage of AI is playing an increasing role in the tragedy that is climate change, and this problem will only worsen as this technology continues to proliferate through society

Nic Lane

Image via www.vpnsrus.com

Machine Learning & Artificial Intelligence

̽��ֱ��text in this work is licensed under a Creative Commons Attribution 4.0 International License. Images, including our videos, are Copyright © ̽��ֱ�� of Cambridge and licensors/contributors as identified. All rights reserved. We make our image and video content available in a number of ways – as here, on our main website under its Terms and conditions, and on a range of channels including social media that permit your use and sharing of our content under their respective Terms.

Yes

Licence type:

Attribution