Where are data and a data centers a matter of life and death?

When a website of a small online store does not open, it causes annoyance, but usually nothing more. However, there are some areas where even a small glitch is a big problem and where the flow of data is critical. For some organisations or services, errors result in major losses or, in extreme cases, even lives lost.

Self-driving car. Photo: TalTech

How much data can be found behind everyday activities?

Let us begin from a little further back, with everyday issues. One of the simplest activities in our lives is sitting in front of a screen for entertainment. ‘For example, a lot of information is exchanged between the device of a customer and the servers of Elisa for the customer to watch a TV show on Elisa Huub,’ explains Joosep Põllumäe, Head of IT and TV Service Development of Elisa.

So, what actually happens when you want to stream your favourite TV show? ‘The device of the customer and its ability to unpack different video streams plays an important role. For this purpose, the device will introduce itself to the servers, and a stream suitable for the specific device is found on the background. It is also important to check whether the logged-in user has the rights to the desired content, i.e. to the appropriate service. In order to do all this, the data of both the device and the user must be moved between different systems,’ explains Põllumäe.

In addition to reliability, secure data storage is also important for streaming. The information must be exchanged in encrypted form and may only be accessible to authorised processors. There are also areas, however, where security and reliability are literally vital.

Will a self-driving car drive off the road if data traffic is interrupted?

The issue of data and self-driving vehicles is a bit more complex, as a lot depends on the type and functionality of the self-driving vehicles observed. According to Raivo Selli, Programme Director of Product Development and Robotics and Professor of Robotics at the School of Engineering of TalTech, the question is not only technological, but also legal, i.e. who is responsible – man or machine (artificial intelligence).

‘The operating principles of self-driving vehicles can vary greatly depending on their purpose and application,’ Sell explains. Modern passenger cars (Level 2) use advanced driver assistance systems (ADAS), but the responsibility lies entirely with the human driver. In most cases, camera and radar technology is used in such vehicles to position the vehicle on the road and with respect to other vehicles. Typically, real-time analysis of road markings and image processing of the road ahead and other objects is performed. This is used to create situational awareness and to control the acceleration and turning of the vehicle. Generally, the human driver will be asked to confirm the turn (the same car may behave differently in Europe and America, depending on the legislation).

The most common Level 4 self-driving vehicles are slow-moving minibuses that operate in a specific area or route and are equipped with pre-recorded information specific to that area. Cameras and LiDAR laser scanning devices are usually used. ‘The vehicle positions itself by comparing the pre-recorded LiDAR map with the real-time LiDAR map. Cameras are, above all, used for detecting objects and are not directly needed for making driving decisions. Such vehicles also do not observe pavement markings, but rather hold this information in their memory as a vector map,’ explains the robotics professor.

It is therefore important that the data on the pre-recorded cards is correct and protected from external manipulation. ‘If the original map (essentially a point cloud) is corrupted, this may result in the vehicle losing its localisation during operating and believing that it is, for example, in the oncoming lane, even though this is not actually the case. As a result, it will start to move back into its own lane, but may actually run off the road,’ says Sell. The point cloud pre-recorded with LiDARs for bus navigation is therefore a critical data set that must be complete and protected.

A ‘vector map’ may also be used to define virtual road markings, lanes, pedestrian crossings, and other means of traffic management. ‘If a vector map has been maliciously modified by cyber criminals, it may result in a dangerous situation on the road or, in the best case, impede the operation of the vehicle,’ notes Sell.

This is why it is first necessary to contribute to cyber security. This means that the control software of a vehicle must be built to protect critical data and processes from cyber attacks. ‘Furthermore, there should be an active attack detection process that monitors unusual network traffic, data integrity, and so on,’ says the robotics professor.

On top of all this, a self-driving vehicle must have the basic capability to manoeuvre safely to the side of the road and drop people off without external assistance in the event of an attack or a threat to data integrity.

The more the movements and decisions of the machine depend on remote IT devices or, for example, on the interaction between different vehicles in free-flowing traffic (e.g. positioning with respect to one another), the higher the importance of the smooth flow of data.

Medicine is evolving towards ‘remotely controlled’ surgeries

So-called remotely controlled robotic medical systems enable to carry out certain procedures, such as surgeries, treatment procedures, and diagnosing over shorter or longer distances by using wired and/or wireless communication networks.

The first spinal robot arrived in Estonia, at the Centre of Spinal Surgery of the East Tallinn Central Hospital in the autumn of 2019, and Rasmus Allikvee, spinal surgeon of the Confido Medical Centre, was one of the first in Estonia and in the neighbouring countries to perform a robotic surgery, Confido reported. ‘At the time, Estonia was the fifth country in Europe to get such a novel device, and it was a very significant leap forward for us in the treatment of spinal problems. ‘Robotic surgeries are much more precise and quicker than open surgeries, and patients also recover faster,’ noted Allikvee.

What kind of data is required for such a surgery? According to the surgeon, the patient will first undergo a computerised examination and the surgery will be planned on the screen. ‘The picture shows exactly at what angle and where the screws should be placed to get the most out of them. During the surgery, the robot helps to ensure that the implants are in the right position and reach the right place,’ explained Allikvee.

Speaking about the future of spinal surgery, Allikvee said the role of robots is increasing. ‘Medicine has developed incredibly rapidly in recent years and I do not see open surgeries ever returning. On the contrary – robots are becoming more precise and skilful, and their use is increasingly more accessible and reasonable. Robots are already involved in a third of all surgeries and their role will only grow,’ the surgeon predicted.

It seems obvious that the movement of data and the reliability of the IT equipment involved in moving the data must be impeccable. Any error would come at a very high price.

So what will ultimately ensure reliable data traffic?

In this regard, attention needs to be paid to the aforementioned software solutions, cyber security, and communication channels, but also to IT equipment and the supporting infrastructure. When it comes to the latter, the demand is covered by the data center of Greenergy Data Centers on the outskirts of Tallinn, which was primarily built to ensure the highest reliability and security. ‘We are also the only large-scale data reception, storage, and transmission complex in the Baltics. The reliability and security of the digital infrastructure is our main goal and our daily work,’ explains Tõnu Grünberg, CEO of Greenergy Data Centers.

Conclusively, the more critical the data, the better idea it is to ask your service providers what would happen in the case of unexpected events or disasters. Is there a Plan B behind Plan A, and preferably also a Plan C? While we can survive a couple of hours without Netflix, we would not want to experience such communication gaps behind the wheel of a self-driving car or on an operating table.