Connect with us

Technology

‘Europe has fallen behind in AI commercialisation’

Voice Of EU

Published

on

AI expert Dr Feiyu Xu talks about the different approaches to AI globally and how natural language processing has changed throughout her career.

Click to read more stories from Automation Week.

A big part of automation is the use of machine learning and artificial intelligence. However, the ways in which these technologies are deployed depend on many external factors, from funding and investment to regulations and location.

Future Human

Dr Feiyu Xu, the global head of AI at German software giant SAP, has a unique view on this due to her background.

Having grown up in China, Xu completed her undergraduate, master’s degree and doctoral studies in artificial intelligence in Germany. She then began her career as a scientist and worked for many years in research in artificial intelligence.

She worked at the German research centre for AI (DFKI) and co-founded and managed an AI start-up before moving into industry.

“First I went to Lenovo, which took me back to my home country, China,” she told SiliconRepublic.com.

“My recent stay in China made me realise how strongly China embraces AI because the need for automation and intelligence in their civil infrastructure is so urgent. In a country with 1.4 billion inhabitants, innovation, in particular, big data and AI technologies are needed to improve the standard of life and work.”

Her view of both Eastern and western cultures has given Xu a unique insight into how AI is being used across the world. She said there are currently at least three approaches to AI globally.

As mentioned, she said the Chinese or Asian way, tends to be very open to the use of big data and AI and where the state invests massively in digital solutions. “In particular, the commercialisation of AI applications has been very successful. “

In the US, Xu said AI innovation is led by large corporations and enabled by their investments. “The US is leading the AI technology research and AI applications.”

Finally, the European approach she said is often focused on regulation and safe-guarding before innovation and where she said “public opinion is still rather sceptical about digital transformation, AI, and big data”.

“Europe has been very successful in basic research and also has a long tradition in AI research. But when it comes to commercialising AI, European industry has fallen behind the US and China, especially in AI for the internet and consumer products.”

Xu said this is clear from the book AI Superpowers by Kai-Fu Lee, where the author sees Chin and the US as the superpowers, while Europe isn’t even a close third place. Additionally, a Deloitte study said that in Germany, companies favour buying off-the-shelf AI rather than developing it themselves.

‘The stricter regulations [in Europe] force us to develop rules and methods to deal with the challenges’
– DR FEIYU XU

Xu said there is a realistic chance for Germany to become a leader in the international AI race if it capitalises on its ability to develop AI, especially in the enterprise software arena. “For Europe, I see increased opportunities in the field of business AI, such as enterprise AI, industrial robotics, health AI and smart manufacturing.”

This is not the first time Europe has been called out for lagging behind other nations.

Earlier this year, a report from the European Parliament’s special committee on artificial intelligence in a digital age said that the EU had “fallen behind” in the global tech leadership race.

“We neither take the lead in development, research or investment in AI,” the text stated. “If we do not set clear standards for the human-centred approach to AI that is based on our core European ethical standards and democratic values, they will be determined elsewhere.”

The lag in innovation is believed to be partially due to the level of regulations around AI technology in the EU. In April 2021, the European Commission proposed new standards to regulate AI in a bid to create what it calls “trustworthy AI”. These proposals seek to classify different AI applications depending on their level of risk and implement varying degrees of restrictions.

However, Xu said that while the legal frameworks in Europe appear “seem very strict,” she said there are ways the EU can turn this into an advantage.

“The stricter regulations force us to develop rules and methods to deal with the challenges. The GDPR and the emerging AI regulations require the explainability and transparency of AI solutions that contribute to decision-making,” she said.

“On the one hand, they pose more hurdles for AI development. Thus, they urge AI research and development to invest more effort in trustworthy AI.”

How natural language processing has changed

A major area of Xu’s expertise lies in natural language processing (NPL), which is a computer program’s ability to understand human language, whether it be written or spoken.

In 2013, Xu won a Google Focused Research Award for her contribution in the field of NLP. She said the pace at which NPL has advanced in recent years is “truly unprecedented”, with many of previously deemed unsolvable problems having since been solved.

“Looking at recent high-profile results like PaLM in which pre-trained models explain common sense reasoning (and explain why jokes are funny), or DALL-E generating images from textual descriptions, the boundaries have yet to be established,” she said.

“I am most excited about the fact that these advances also have a major impact on business AI, as many of the advances are about getting done more but with less data – and access to data is always an obstacle to applying AI in the enterprise.”

She said that at the start of her research career, working in NLP meant applying a variety of means, ranging from rule-based methods for basic tasks to statistical measures and graph algorithms, all the way to traditional machine learning.

‘With each leap, NLP is producing better results with fewer data points’
– FEIYU XU

“Each problem was addressed by a specific combination of these methods, and each NLP researcher needed a deep understanding of each of those to develop solutions,” she said.

“With the advent of deep learning methods, NLP solutions started to look more similar. Early on, deep learning was considered yet another tool in the box, but as it significantly increased accuracy on many tasks, it was used more and more.”

These advances then led to the emergence of transformer-based pre-trained language models such as BERT and GPT-2. These models were trained on a vast number of texts by trying to complete sentences or fill in blanks and the focus for solving NLP tasks switched from methods to data.

“The most recent leap, where bigger and bigger models, based on the same transformer components as BERT, are trained on more and more data, enables these models [such as] GPT-3 to address NLP tasks without even fine-tuning,” she said. “The models auto-complete the next examples by simple pattern matching, with surprisingly sophisticated and usable results.

“With each leap, NLP is becoming easier to apply to new tasks, requiring less knowledge, and producing better results with fewer data points.”

Beyond NLP, Xu said there are two AI trends she sees having a major impact in the future, the integration of information extracted from texts and from structured sources such as data bases, and the explainability of black-box machine learning.

She said the information integration will “enable the explicit representation of knowledge and enable machines and humans to work on structured knowledge jointly”, which will be crucial for business AI where “correctness is paramount”.

In terms of black-box machine learning methods, she said transparency will be key to the success of business AI.

When enterprise users work with machine learning-based recommendations or predictions, users need to understand how they came to be able to judge if they can be trusted to identify errors and mistakes,” she said.

“With transparency then, machine learning methods can simplify the lives of enterprise users, allowing them to get their work done more quickly, and plan their businesses with greater foresight.”

10 things you need to know direct to your inbox every weekday. Sign up for the Daily Brief, Silicon Republic’s digest of essential sci-tech news.

Source link

Technology

EU-backed project to trial uncrewed flight ecosystem in Shannon

Voice Of EU

Published

on

The Shannon-based project aims to integrate the operations of uncrewed and conventional aircraft to modernise air traffic management in Europe.

A European consortium based in Shannon has received EU funding to develop a flight ecosystem for drones and help integrate uncrewed aircraft into our airspace.

Coordinated by Future Mobility Campus Ireland (FMCI), this consortium will conduct a three-year engineering project to develop, deploy and optimise this type of system in Europe.

Describing itself as Ireland’s “first testbed for future mobility”, FMCI is a development centre based in the Shannon Free Zone focused on innovation in both ground and air mobility tech.

Illustration of an unmanned vehicle testing site, with drones visible. A landing and take off zone is highlighted, along with a mobile operations unit where a van is parked. A small building is labelled as the AAM operations centre.

Illustration of the Advanced Aerial Mobility Hub at FMCI. Image: FMCI

FMCI said the research project, known as EALU-AER, represents a “major vote of confidence” in Ireland’s local expertise, industry operators and the resourcing of air mobility development.

Other members of the consortium include Shannon Group, the Irish Aviation Authority, Collins Aerospace, Dublin-based Avtrain, and Deep Blue in Italy.

The consortium has received the three-year funding award to develop uncrewed aviation business opportunities in Ireland, as part of a collaborative research project that could help modernise air traffic management in Europe.

The consortium said the new funding will help build an end-to-end ecosystem that supports the safe operation of uncrewed flights. The goal is to help integrate the operations of both uncrewed and conventional aircraft.

“This will result in developing and building out the critical infrastructure to allow advanced air mobility proliferate across Europe,” FMCI CEO Russell Vickers said.

“It will secure access to airspace for large numbers of drones and eVTOL [electric vertical take-off and landing] aircraft, resulting in safe, cost-effective and sustainable transport of freight and people in the future.”

The project’s work will be based at FMCI’s Advanced Aerial Mobility Research Test and Development Facilities in Shannon, but will include a network of Advanced Air Mobility routes across Ireland.

FMCI has already worked with Avtrain and Shannon Group to trial freight delivery services using beyond visual line of sight (BVLOS) drones.

“We are entering a new era of innovation where the success of the industry will depend on the integration of uncrewed aircraft into our airspace, rather than the segregation of airspace,” Avtrain CEO Julie Garland said.

Funding for the project came from the SESAR 3 Joint Undertaking, which is partnership of private and public sector entities in the EU that aim to accelerate the delivery of the Digital European Sky through research and innovation.

It comes as people are increasingly looking at the potential of drones and uncrewed flight technology. A Dublin City Council initiative recently looked to show how local government can utilise drones in areas such as civil defence, emergency response, public safety and environmental monitoring.

10 things you need to know direct to your inbox every weekday. Sign up for the Daily Brief, Silicon Republic’s digest of essential sci-tech news.

Source link

Continue Reading

Technology

Goodbye silicone? A new era of breast reconstruction is on the horizon | Breast cancer

Voice Of EU

Published

on

Having an ice pack strapped to your chest – that’s how some describe the experience of taking a walk in cold weather when you have breast implants. Silicone only slowly reaches body temperature once out of the cold, so that icy feeling can persist for hours. As well as being uncomfortable, for breast cancer survivors it can be an unwelcome reminder of a disease they would rather put behind them.

Every year, 2 million people worldwide are diagnosed with breast cancer and the treatment often involves removing at least one breast. But most choose not to have their breasts reconstructed; in the UK, it is only about 30%. Now a handful of startups want to change that, armed with 3D-printed implants that grow new breast tissue before breaking down without a trace. “The whole implant is fully degradable,” says Julien Payen, CEO of the startup Lattice Medical, “so after 18 months you don’t have any product in your body.”

It could spell the end not only of cold breasts, but the high complication rates and long surgeries associated with conventional breast reconstruction. The first human trial of such an implant, Lattice Medical’s Mattisse implant, is scheduled to begin on 11 July in Georgia. Others will soon follow. “We expect to start clinical trials in two years’ time,” says Sophie Brac de la Perrière, CEO of another startup, Healshape.

“It’s exciting,” says Stephanie Willerth, professor of biomedical engineering at the University of Victoria, Canada, who is not involved with the companies. “As engineers, we’ve been playing with 3D printing for half a decade”, but having a clinical use that doctors recognise as useful for patients is key to getting the technology out there, she says.

But in a field fraught with difficult medical compromises, unequal access issues and expectations about what women want, the question is how big an impact the new technology will actually have.


Today, there are two main types of breast reconstruction: silicone implants and flap surgery. While implants are easy to install, flap surgery is a highly specialised business that requires a tissue “flap” being taken from the stomach, thigh or back. Surgeons often recommend flaps because, while there’s a lot of initial surgery and a longer recovery period, it gives a good, long-lasting result.

Silicone is still the most common choice. It is easy and simple, which appeals to cancer patients who either medically can’t have or mentally can’t face having tissue removed from another part of their body. But “it’s far from perfect”, says Shelley Potter, an oncoplastic surgeon at the University of Bristol and the Bristol Breast Care Centre. “It’s quite high risk. There’s a 10% chance of losing an implant.”

Healshape’s 3D-printed hydrogel implant
Healshape’s 3D-printed hydrogel implant, designed to be colonised by the patient’s fat cells over six to nine months. The company hopes to start trials in two years’ time. Photograph: Healshape

Silicone implants also require replacement every 10 or so years and they have had their fair share of scandals: the 2010s PIP scandal, in which a major implant manufacturer was found to have made its implants of dodgy silicone, and the 2018 Allergan scandal, in which popular textured implants were linked to an increased risk of a rare lymphoma. And as an American study from last year shows, it is mainly the idea of having that foreign object stuck inside your body that puts many off reconstruction altogether.

“So what we want to do,” says Brac de la Perrière, “is to give the benefits of the different solutions without the constraints.” In other words: the single, simple surgery of an implant, but without any lingering foreign material to cause trouble.

This can be achieved in different ways. Healshape uses a hydrogel to 3D-print a soft implant that will slowly be colonised by the person’s own fat cells, the initial batch of which is injected, while the implant disappears over six to nine months. The company CollPlant is developing something similar using a special collagen bioink, extracted from tobacco leaves it has genetically engineered to produce human collagen. “I think it will change the opinion of many patients,” says CEO, Yehiel Tal.

Lattice Medical has a different approach. Its implant is a 3D-printed cage made of a degradable biopolymer, in which they encase a small flap from underneath the breast area. This flap then grows to fill the cage with fat tissue, while the cage itself is absorbed by the body, ultimately leaving a regrown breast in its place.

Lattice Medical’s Mattisse implant
Lattice Medical’s Mattisse implant. Vascular adipose tissue is inserted into a bio-resorbable ‘tissue engineering chamber’, which degrades over 18 months. Trials are imminent. Photograph: Lattice Medical

Regrowing breasts using a cage has been shown to work in humans before, in a 2016 trial. However, it only worked in one of five women and the cages were not degradable. Andrea O’Connor from the University of Melbourne, Australia, who led the trial’s engineering team, hopes the new trial will address the problems raised in the first – for example, that patient responses can vary greatly. But if successful, it “would have the potential to help many women to achieve a superior reconstruction”, she says. Lattice Medical says its cage is an improvement because a flat base and larger pores help the tissue grow.

One big unknown is how much feeling the regrown breasts will have. A mastectomy usually means losing some sensation and, according to plastic surgeon Stefania Tuinder from the Maastricht University Medical Centre+ in the Netherlands, reconstruction affects it too. “From our data, it seems that implants have a negative effect on sensation, so the feeling in the skin is less than when you have only a mastectomy,” she says. In comparison, reconstruction from a flap with connected nerves can bring back some feeling within a few years.

Tuinder suspects the implant numbness is both because of nerve damage when the implants are inserted, and because the nerves can’t grow back once they are blocked by a lump of silicone. Whether that will also apply to the new implants remains to be seen, but since eventually there will be nothing to block the nerves, hopes are that sensation will be better.


Tissue engineered implants, however, are not the only recent innovations in the field. Many groups are working on perfecting a reconstruction technique using injections of the person’s own fat, boosted with extra stem cells to help the tissue survive. Medical professionals are still debating the safety and how the breasts hold up long term. In contrast to the new implants, the procedure might have to be done several times.

While any of these new techniques could result in something better than what’s currently on offer, Potter warns that we have a tendency to jump at new and shiny tech – an optimism bias. “We always think it’s going to be brilliant,” she says, but “we don’t want a situation like with vaginal mesh, where in 10 years’ time … we find out we have done something that isn’t helpful.”

Other solutions to the problems of reconstruction do exist. One is living without breasts, known as “going flat”. Contrary to the companies that think they can turn the reconstruction statistics around, people within the flat movement argue that if people were better informed, even more would opt out. “I reckon if [going flat] was given as an equal option,” says Gilly Cant, founder of the charity Flat Friends, “at least another 30-50% of women wouldn’t have [reconstruction].”

A Healshape scientist using software to determine the shape of an implant prior to 3D printing. The implants can be custom-made to suit the patient.
A Healshape scientist using software to determine the shape of an implant prior to 3D printing. The implants can be custom-made to suit the patient. Photograph: Healshape

At the moment, the guidance from the National Institute for Health and Care Excellence (Nice) says that doctors should be aware that some might not want reconstruction. But Cant says it is often presented to people as part of the treatment process. “It’s like, ‘OK, we need to do a mastectomy. Then you have chemo. Then you’ll have your radiotherapy and then we’ll do reconstruction.’ So women live for that reconstruction at the end,” she says. It comes to signal the finish line.

It is particularly contentious when only one breast is removed, because some might want the other taken off to feel and look symmetrical, rather than have a new one made. But according to Cant, many doctors don’t want to remove a healthy breast. Part of the doctors’ concern is that women will regret their decision, says Potter, but “women know what they want to do with their own bodies. We should help and support them to do what they want to do.”

Potter herself would like to see more of the ultimate alternative: not having a mastectomy in the first place. “There’s no evidence that mastectomy gives you better cancer outcomes than a breast-conserving operation,” she says. In this case, the tumour is removed but the breast is kept. For example, one of her patients had a breast reduction that removed her cancer while giving her breasts a lift. “She calls them her silver lining breasts.”


So even without tissue-engineered implants, there are enough options to make the choice a hard one. To help people choose, some charities pair up people considering a specific procedure with someone who has already been through it. At the charity Keeping Abreast, show and tell sessions give people the chance to ask the questions they might be uncomfortable asking their doctor and see the results for themselves.

But according to a 2018 report by the all-party parliamentary group on breast cancer, knowing what you want is not the same as having access to it. “There’s a massive postcode lottery,” says Potter. It stems from flap surgery being so involved that it often requires specialist plastic surgeons who can do minute surgery under a microscope. Many clinics don’t have such experts in-house and while the Nice guidance says people should still have the option, in practice it limits access.

The companies say this won’t be a problem with the new implants, because they are specifically designed to be easy to put in. Flap surgery can take from three to 12 hours depending on the flap, but insertion of Lattice Medical’s implant, for example, takes only one hour and 15 minutes. “It’s really accessible to all plastic surgeons,” says Payen.

This accessibility will no doubt be crucial in taking the new implants from a cool technology to something with real impact. But from Potter’s perspective, it’s just one potential piece in a big puzzle, not a techno-fix. The implants “would be an option for a lot of women”, she says. “But I think the main advance is all around access, proper information, giving women choice and hopefully reducing the number of mastectomies that we need.”

Source link

Continue Reading

Technology

What to do about inherent security flaws in ICS? • The Register

Voice Of EU

Published

on

The latest threat security research into operational technology (OT) and industrial systems identified a bunch of issues — 56 to be exact — that criminals could use to launch cyberattacks against critical infrastructure. 

But many of them are unfixable, due to insecure protocols and architectural designs. And this highlights a larger security problem with devices that control electric grids and keep clean water flowing through faucets, according to some industrial cybersecurity experts.

“Industrial control systems have these inherent vulnerabilities,” Ron Fabela, CTO of OT cybersecurity firm SynSaber told The Register. “That’s just the way they were designed. They don’t have patches in the traditional sense like, oh, Windows has a vulnerability, apply this KB.”

In research published last week, Forescout’s Vedere Labs detailed 56 bugs in devices built by ten vendors and collectively named the security flaws OT:ICEFALL. 

As the report authors acknowledged, many of these holes are a result of OT products’ being built with no basic security controls. Indeed, Forescout’s analysis comes ten years after Digital Bond’s Project Basecamp that also looked at OT devices and protocols and deemed them “insecure by design.”

A few hours after Forescout published its research, CISA issued its own security warnings related to the OT:ICEFALL vulnerabilities.

CVEs: The problem? Or the fix?

“Up until this point, CVEs haven’t been generated for these insecure-by-design-things, and there’s a reason for that,” Fabela said. “It’s bad for the industry.”

Once a CVE is generated, it sets into motion a series of actions by industrial systems’ operators, especially in heavily regulated industries like electric utilities and oil and gas pipelines. 

First, they have to determine if the environment contains any affected products. But unlike enterprise IT, which usually has centralized visibility and control over IT assets, in OT environments, “everything is distributed,” Fabela noted.

If industrial and manufacturing environments do have any products impacted by the vulnerability, that triggers an internal review and regulatory process that involves responding to CISA and developing a plan to improve security.

One SynSaber customer sarcastically described OT:ICEFALL as “the gift that keeps on giving,” Fabela said. “He said, ‘Now I have this on top of all my other like, the real vulnerabilities’,” which present a slew of other problems when it comes to patching — such as having to wait until a planned maintenance outage that may be months out — if the manufacturer has a patch at all.

OT protocols don’t use authentication

For example: The current Modbus protocol, which is very commonly used in industrial environments, does not have authentication. 

Forescout’s analysis details nine vulnerabilities related to unauthenticated protocols and disputes the argument that against assigning a CVE ID to a product with an insecurity OT protocol.

“On the contrary, we believe a CVE is a community recognized marker that aids in vulnerability visibility and actionability by helping push vendors to fix issues and asset owners to assess risks and apply patches,” the authors wrote.

While this makes sense from an IT security perspective, Fabela said it’s unrealistic from an OT perspective, and ultimately doesn’t make critical infrastructure any more secure.

Modbus, as a protocol that does not use authentication, could generate “thousands” of CVEs that “affect every product line in the world,” he Fabela. “You’re tying up the product security teams with the OEMs and you’re tying up the customers, the asset owners with CVE that they can’t do anything about.” 

Basecamp researcher weighs in

Reid Wightman is a senior vulnerability researcher with OT security shop Dragos’ threat intel team. He’s also one of the original Project Basecamp researchers, and, more recently has done work on the ProConOs and MultiProg software vulnerabilities.

Forescout cited some of his research, and dedicated a section of the ICEFALL analysis to security flaws with the ProConOS runtime in PLCs.

In an email to The Register, Wightman noted that a lot of industrial controllers have the same set of problems that isn’t going away: “they allow unauthenticated code to run on the PLC.” 

“This means that one malicious logic transfer to the PLC may permanently compromise the PLC,” he added, noting that, because the control logic is causing the change, it can happen outside of a normal firmware update. “It’s kind of a thing I’ve harped on since the Basecamp days, but may be worth repeating. Over and over again. Until the sun burns out, probably.”

Lately, one of Wightman’s “big, personal concerns” is that some vendors say they can use TLS and client certificates to secure controllers, presumably to avoid. In reality, this would just make the traffic more difficult to inspect, Wightman said.

“If an attacker gets onto the engineering system, they may load a malicious payload using CVE-2022-31800/CVE-2022-31801 (or any of the similar problems that exist in almost every logic runtime) into the controller,” he added. “Only, now we have no way of telling whether they did it because the traffic is encrypted.”

So how do we fix the problem? 

“I guess my answer would be: if your engineering system is compromised, throw away all of the controllers that it was allowed to talk to,” Wightman said. “And I doubt most end users would go to that level of paranoia.”

Which, again, points to the insecure-by-design nature of how these systems are engineered.

“Thankfully, we see no signs of any widespread abuse of these protocols or ‘features’ in spite of some of the bugs being well-known for years,” Wightman added. “I really do hope it stays that way.” ®

Source link

Continue Reading

Trending

Subscribe To Our Newsletter

Join our mailing list to receive the latest news and updates 
directly on your inbox.

You have Successfully Subscribed!