What’s your Face Worth?

AI and Facial Recogntion

While most European citizens remain wary of AI and Facial Recognition, Maltese citizens do not seem to grasp the repercussions of such technology. Artificial Intelligence expert, Prof. Alexiei Dingli (University of Malta), returns to THINK to share his insights.

The camera sweeps across a crowd of people, locates the face of a possible suspect, isolates, and analyses it. Within seconds the police apprehend the suspect through the capricious powers of Facial Recognition technology and Artificial Intelligence (AI).

A recent survey by the European Union’s agency for fundamental rights revealed how European citizens felt about this technology. Half of the Maltese population would be willing to share their facial image with a public entity, which is surprising given that on average only 17% of Europeans felt comfortable with this practice. Is there a reason for Malta’s disproportionate performance? Artificial Intelligence expert, Prof. Alexiei Dingli (University of Malta), returns to THINK to share his insights.

Facial Recognition uses biometric data to map people’s faces from a photograph or video (biometric data is human characteristics such as fingerprints, gait, voice, and facial patterns). AI is then used to match that data to the right person by comparing it to a database. The technology is now advanced enough to scan a large gathering to identify suspects against police department records. 

Data is the new Oil

Facial Recognition and AI have countless uses. They could help prevent crime and find missing persons. They are prepared to unlock your phone, analyse, and influence our consumption habits, even track attendance in schools to ensure children are safe. But shouldn’t there be a limit? Do people really want their faces used by advertisers? Or, by the government to know about your flirtation with an opposing political party? In essence, by giving up this information, will our lives become better?  

‘Legislation demands that you are informed,’ points out Dingli. Biometric data can identify you, meaning that it falls under GDPR. People cannot snap pictures of others without their consent; private data cannot be used without permission. Dingli goes on to explain that ‘while shops are using it [Facial Recognition Technology] for security purposes, we have to ask whether this data can lead to further abuses. You should be informed that your data is being collected, why it is being collected, and whether you consent or not. Everyone has a right to privacy.’

Large corporations rely on their audiences’ data. They tailor their ad campaign based on this data to maximise sales. Marketers need this data, from your Facebook interests to tracking cookies on websites. ‘It’s no surprise then,’ laughs Dingli, ‘that Data is the new oil.’ 

The EU’s survey also found that participants are less inclined to share their data with private companies rather than government entities. Dingli speculates that ‘a government is something which we elect, this tends to give it more credibility than say a private company. The Facebook-Cambridge Analytica data breach scandal of 2018 is another possible variable.’ 

China has embraced Facial Recognition far more than the Western World. Millions of cameras are used to establish an individual citizens’ ‘social score’. If someone litters, their score is reduced. The practise is controversial and raises the issue of errors. Algorithms can mismatch one citizen for another. While an error rate in single digits might not seem like a large margin, even a measly 1% error rate can prove catastrophic for mismatched individuals. A hypothetical 1% error rate in China, with a population of over 1.3 billion, would mean that well over ten million Chinese citizens have been mismatched.   

Is privacy necessary?

‘I am convinced that we do not understand our rights,’ Prof. Dingli asserts. ‘We do not really value our privacy and we find it easy to share our data.’ Social media platforms like Facebook made its way into our daily lives without people understanding how it works. The same can be said for AI and facial recognition. It has already seeped its way into our lives, and many of us are already using it—completely unaware. But the question is, how can we guarantee that AI is designed and used responsibly?

Dingli smiles, ‘How can you guarantee that a knife is used responsibly? AI, just like knives, are used by everybody. The problem is that many of us don’t even know we are using AI. We need to educate people. Currently, our knowledge of AI is formed through Hollywood movies. All it takes is a bit more awareness for people to realise that they are using AI right here and now.’

Everyone has a right to privacy and corporations are morally bound to respect that right, individuals are also responsible for the way they treat their own data. A knife, just like data, is a tool. It can be used for both good and evil things. We are responsible for how we use these tools. 

To Regulate or Not to Regulate?

Our data might not be tangible, but it is a highly valued commodity. Careless handling of our data, either through cyberattacks or our own inattention, can lead to identity theft. While the technology behind AI and Facial Recognition is highly advanced, it is far from perfect and is still prone to error. The misuse of AI can endanger human rights by manipulating groups of people through the dissemination of disinformation.   

Regulating AI is one possibility; it would establish technical standards and could protect consumers, however, this may stifle research. Given that AI is a horizontal field of study, fields such as architecture and medicine must consider the implications of a future with restricted use. An alternative to regulation is the creation of ethical frameworks which would enable researchers to continue expanding AI’s capabilities within moral boundaries. These boundaries would include respecting the rights of participants and drawing a line at research that could be used to cause physical or emotional harm or damage to property. 

While the debate regarding regulation rages on, we need to take a closer look at things within our control. While we cannot control where AI and Facial Recognition technology will take us, we can control whom we share our data with. Will we entrust it to an ethical source who will use it to better humanity, or the unscrupulous whose only concern is profit? 

Further Reading:

The Facebook-Cambridge Analytica data breach involved millions of Facebook users’ data being harvested without their consent by Cambridge Analytica which was later used for political advertising; 

Chan, R. (2019). The Cambridge Analytica whistleblower explains how the firm used Facebook data to sway elections. Business Insider. Retrieved 8 July 2020, from https://www.businessinsider.com/cambridge-analytica-whistleblower-christopher-wylie-facebook-data-2019-10.

Malta’s Ethical AI Framework;Parliamentary Secretariat For Financial Services, Digital Economy and Innovation. (2019). Malta Towards Trustworthy AI. Malta’s Ethical AI Framework. Malta.AI. Retrieved 8 July 2020, from https://malta.ai/wp-content/uploads/2019/10/Malta_Towards_Ethical_and_Trustworthy_AI_vFINAL.pdf

Let’s rethink our motors

Daniel Lendi debugging HDMS

From washing machines and vacuum cleaners to air conditioners and fans, household appliances all contain electrical motors, and we tend to take them for granted. However, appliances’ high starting currents can disrupt electricity supply and shut down a whole house. Daniel Lendi unpacks some of the hard work needed to fix this problem.

My (Daniel Lendi) tech-savvy generation takes many things for granted. Entering the Electrical and Electronics Engineering course at the University of Malta (UM) helped me realise just how much thought and effort humanity has invested in technology. My passion for electrical motors and their applications grew further at university, and until today, I’m fascinated by how an electrical current passing through some copper wire can lead to the physical rotation of a motor shaft, driving electric cars and pumping water. 

My time at university also taught me that these motors introduce complications when they start. High currents can trip the Miniature Circuit Breakers inside your consumer unit. When ACs start they draw six to eight times their normal amount of energy; switching on your AC could even make your neighbour’s lights flicker! Dedicated motor starters are introduced to tackle these complications. Dr Ing. Reiko Raute (Department of Industrial Electrical Power Conversion, Faculty of Engineering, UM), with whom I worked on my dissertation, has come up with a novel prototype that aims to improve on these motor starters. Being part of his team allows me to further my studies while gaining valuable insight into industrial product development.

Justin Xuereb, Dr Ing. Reiko Raute and Daniel Lendi
Justin Xuereb, Dr Ing. Reiko Raute and Daniel Lendi. Photo by Racquel Ellul

One type of these motor starters is the soft starter (a category of reduced voltage starters). The soft starter decreases the voltage to the motor, reducing its current and keeping your house lit up. However, as a consequence, they reduce the motor’s torque – a measure of the motor’s rotational strength. A lower torque means that the motor doesn’t turn strongly enough, making your AC inefficient and possibly damaging the motor.

Raute is trying to overcome these problems with a new prototype High Dynamic Motor Starter (HDMS). The HDMS is a soft starter which tackles the issue through a fundamentally different approach. Using a dedicated control algorithm (computer code that carries out specific tasks) and a novel hardware circuit, the HDMS is able to achieve excellent starting current reduction without sacrificing so much motor torque. The development of this prototype is part of a research collaboration between UM and the development division of electronic equipment designer and manufacturer Carlo Gavazzi. Our industrial partner ensures that the product meets market expectations. Prototypes of the HDMS soft starter have already been constructed and are now being tested in heat pump installations in the UK, France, Ireland, and the USA, collecting data from each motor start they carry out. 

The development of the new starter is an ambitious feat. Our team divided it into two projects. My part is analysing the HDMS control algorithm stability and its optimisation. I took up a study module on control engineering to better understand what’s going on. The aim is to ensure that the HDMS regulates and reduces the starting current in the most stable and efficient way, using both the simulation models and the hardware prototype. For the hardware tests we built a test rig which is able to measure a motor’s torque during start-up. I have concluded the analysis, the most tedious part of my project, and am finally moving on to hardware tests to confirm its findings.

Our industrial partner ensures that the product meets market expectations.

My colleague Justin Xuereb focuses on the second project. He investigates the dynamic overload performance of the switching devices. Think of dynamic overload performance as similar to a workout: You are at the gym training with some weights, and are capable of carrying out your routine with 40kg of weights. But you want a bigger challenge and add another 20kg of weights to see how you perform for a few repetitions. This is what is done with the dynamic overload performance of the switching devices, overloading them for a short amount of time and studying their performance. Xuereb has already designed and implemented several revisions of a dedicated test circuit to investigate the overload performance of the switching devices.

My analysis has shown that Raute’s creative idea adheres to control engineering theory. The work is still continuing to ensure the HDMS’ market success. When the next-generation single-phase soft starter product for Carlo Gavazzi becomes a marketable product, our ACs will start more efficiently, even in the midst of summer.   

The project is funded by the Malta Council for Science and Technology (MCST) R&I Fusion Technology Development Project R&I-2016-035-T.

Moving energetically ahead

UM's solar panels

Can the university lead the way in shedding a dependence on fossil fuels? Renewable sources cover just over 7% of Malta’s energy needs, which must nearly double by 2030 to implement the draft National Energy and Climate Plan. Daiva Repeckaite traces how photovoltaic modules were rolled out on the campus’s roofs — and finds that to be sustainable, the community should love their ACs a little less.

Continue reading

Living in Maltese

Maltese letters

A new European Commission tool facilitates easy and accurate translation between English and Maltese. Author and translator Mark Vella is tirelessly promoting this translation product. He speaks to Teodor Reljic about how the Maltese language and its literary heritage has defined his life and career from graduation onwards.

Continue reading

Allow yourself to get lost

Datsuzoku app interface

Author: Maria Elena Galea

In the fast-paced world of today, the word ‘slow’ often conjures negative feelings. When on holiday, many people rush attractions, taking snapshots to create proof of their visit for social media. In Malta, popular tourist apps guide visitors to overcrowded beaches, while there are a lot of small villages that tourists never discover. As a passionate traveller, I chose to integrate the concept of slow living into my dissertation and explore how user interface design can help people appreciate the place they are visiting. 

Continue reading

Where Archaeology, Physics, and Artificial Intelligence Meet

Author: Dr Marc Tanti

Ancient Egypt is famous for the mummies of Pharaohs, but did you know that there are many mummified animals? Studying them offers scientists a wealth of knowledge on the method and motivation behind this practice. But mummies are fragile artefacts, and museum curators don’t generally appreciate archaeologists dissecting their specimens. To get around this, X-rays help researchers peek inside the mummies without damaging them.

After finishing my Ph.D. in artificial intelligence, I started working as a research support officer at the University of Malta on a collaborative project with archaeologists at the European Synchrotron Radiation Facility (ESRF), a research institute in France. These archaeologists are studying animal mummies from museum collections, such as the Museum of Grenoble, in order to learn about their structure. This institute is better known for its particle accelerator, which sets electrons flying at nearly the speed of light to understand the shape of drugs and other molecules. So, what’s the link with mummies?

Continue reading