Technology Energy Paradox

Energy has never been used so lavishly as it is today. Esteemed scientists suggest that more energy has been used by humanity during the last century than the whole of the previous history – say, in the last 10,000 years.

This couldn’t be any more relevant now, due to the global COVID-19 pandemic.

The power consumption of the Digital Age is often overlooked and too easily forgotten. Due to millions (those of us lucky enough to do so) now working from home because of the pandemic, we are consuming monster proportions of internet data and electricity.

In return, this is increasing our expectations as well as our energy consumption footprint almost in silence.

Increases in residential demand for electricity have spiked whilst industrial demand has fallen. This is perhaps, to be expected considering many economies across the world have grounded to a halt.

With the industrial demand fallen, this enforces downward pressure on coal, gas and nuclear power to cover the difference. The longer the pandemic goes on, the more detrimental this pattern will become upon our planet.

As a consequence of the efforts to slow the spread of the COVID-19, the share of energy use exposed to containment measures jumped from 5% in mid-March to 50% in mid-April this year.

The restrictions put in place from country to country represent a challenging portmanteau problem of trying to accommodate the supply and demand of energy consumption alongside its distribution, especially with regards to electricity and the equivocal role of data centres.

It is as much a global project management problem, as it is a system design problem.

It is evident, that further macroeconomic problems will affect this and Governments alongside legislatures will have to carefully manage the issues at hand.

I shall not dwell upon this too much but authorities will need to consider prevalent issues such as high costs incurred by peak demands; access to consistent tariffs blocked by individual financial barriers; network deployment issues; maintaining diverse vertical industries as well as leveraging national broadband strategies and regulations.

Of course, the internet has come to our rescue during this difficult time. Yet, many of us may have been experiencing serious data lagging, slow download rates, and network performance differences over the past few weeks. I know I have...

Of course, there are other external factors at play but a majority of these issues occur because the accompanying infrastructures in place are now seriously under strain due to the current global demand.

Whether we like it or not, the demands upon the infrastructures are having a reversal effect upon our planet. Yet, it is important not to forget that this has been going on for some time.

The internet as we know it in the industrialized world is greatly dependent on abundant energy supply, a robust electricity infrastructure, and continuous economic growth.

These three variables are susceptible to constant change and are not to be relied upon if we are to continue using the internet the way we currently do.

Most notably, there have been many changes in public opinion about our “digital footprint” and its adverse effects upon the world.

There have also been many significant adjustments made by global technology coverts who supply the demand. But the issues at hand are still problematic and the COVID-19 pandemic has only exacerbated this.

A Misapprehension of the Internet Age

A misapprehension of the Internet Age is probably why debates over digital energy consumption continue.

Over the last decade, we have seen rapid adoption of smartphones coupled with the equally rapid growth of cloud-based computer services which now accumulate and concentrate much of the internet’s usage.

Such transformation, although significant and disruptive from an industrial and operational level, not to mention its planetary importance, also employ a heavy energy consumption burden.

Importantly, the increasing energy consumption of the internet is not so much due to a growing amount of people using the network, as one would assume.

Rather, it’s caused by a growing energy consumption per internet user. This is because the average internet user mismanages their energy consumption when compared to the extensive (but also damaging) data centres that serve and ensure that the internet is accessible for all devices.

What is more, the time an individual spends surfing the internet increases every year thanks to the steady improvements in computational processing power. Thus, the better the processing power the longer battery life, the less interruption one would have when surfing.

Surprisingly, this did not start with smartphones. Rather, it was with laptops.

Although laptops were built and designed to be “lower in energy consumption”, people soon noticed the advantage a laptop had over smartphones when it came to being online for long periods.

Since the ultimate laptop has a mass of 1 kilo, using Einstein’s famous equation relating mass and energy, E = mc^2, the ultimate laptop has an energy of E = mc^2 = 8.9874 x 10^16 joules.

With this in mind, the ultimate laptop can perform roughly 5.4258 x 10^50 operations per second. Each of these operations is single-bit operations, measured today using FLOPS (floating-point operations per second technique).

These numbers are exciting to the trained eye and a technological marvel that the modern world greatly depends upon.

Yet, as new technological advancements evolve, so do our human drives which often prove to be incompatible in their aims and demands. This conundrum best represents our love-affair relationship with technology and the environment.

It was reported recently that our Global Communications Network consumed 1,815 Terawatt’s per hour (TWh) of electricity in 2012. This corresponds to 8% of global electricity production in the same year (22,740 TWh).

Most of this was due to a boom in laptop production. Seeing the differences at play, technological vendors and suppliers such as Apple or Samsung soon began to engineer new possibilities and “shrink” these advantages into smartphones. Eventually, smartphones became the new laptops.

With additional environmental infrastructure implemented around the world, smartphones were able to consume large portions of internet data from virtually anywhere.

This has become a staple in all our lives, and the benefits are tremendous but subsequently, we have made it exceptionally difficult to live without smartphones. The consequences of this are complex and the public seems to be rather reluctant to compromise or reduce this dependency.

With this in mind, we are on the route towards all electrical consumption of computers, smartphones, tablets, pressure cookers, laptops and other gadgets doubling by 2022 and tripling by 2030.

The infrastructure currently in place would have to incrementally increase its capacity every 2 years or less. The economic demands this would put on our societies, as well as the severe environmental demands this would cause, are exponential.

Although air quality has seen significant improvements and there has been a dramatic fall in Global CO2 emissions throughout the current COVID-19 pandemic, few have failed to understand the consumption trade-off between data and energy.

If anything positive has come out of this pandemic, it is that this has highlighted serious flaws in mankind's inept lack of knowledge in systems design. Although this cannot be fixed overnight, it is indeed a process that must be considered if we are to try and change the modern world for the better.

The Consumption Matrix

Coming back to digital energy consumption. The modern approach towards accessing the internet greatly depends on infrastructures composed of data centres.

The data-centres host “servers” to store all of the internet’s information which allows this to be accessible via tariffs (broadband). Now I must stress, not all of the data accessible on the internet is necessarily hosted on data centres. Private vendors also contribute to the global network.

Over the last 10 years, the internet has evolved and changed. Much of the internet’s traffic flows the same way thanks to digital applications or portals, such as social media.

Such applications are more commonly hosted on data centres. With these advancements, only until recently have we been able to identify how much energy consumption end-user devices (laptops, smartphones. etc) produce or contribute towards when accessing the internet, also known as “data traffic”.

According to the Cisco Annual Internet Report, there will be 5.3 billion global internet users, 3.6 global devices and connections per capita, and an average global fixed broadband speed of 110 Mbps by 2023.

Since the internet’s conception, we have seen data traffic concurrently rise much faster than the number of users using the internet.

However, although data traffic has gradually become more energy-efficient, the internet’s total energy intensity increases and this is the catch. This trend can only be stopped when we limit the demand for every user.

The growing energy use of the internet is often put to one side with vendors arguing that the overall network saves more energy than it consumes.

Many vendors and suppliers put forth cunning “substitution effects” to cater for this, in which online services replace other more energy-intensive activities.

Examples are video-conferencing, which technological vendors claim is supposed to be an alternative for the aeroplane or the car or downloading or streaming digital content.

However, such research has been highlighted as fraudulent, due to the assumptions one has to make when trying to address such a matter. Researchers must ensure they accommodate every possible trade-off and ensure their conversions are accurate.

By today’s standards, researchers and scientists have labelled this methodology a Consumption Matrix.

Furthermore, energy consumption per internet user is increasing with each bit rate of content. The internet started as a text medium, but images, music and video have become just as important. As highlighted by several researchers - “estimates of the energy intensity of the internet diverge by several orders of magnitude”.

If researchers fail to understand how many diverges are actually at play (also known as sensitivity analysis), it becomes incredibly difficult to analyse energy consumption across different sectors.

As digital energy consumption rises with every bit of data, it matters a lot what we’re doing online. As it turns out, we are increasingly using the network for content with high bit rates, especially video and music streaming.

Rise of Wireless Networks

Smartphones have evolved to such an incredible degree that many, if not all, can access the internet wirelessly using 3G or 4G broadband.

By the end of 2020, the number of smartphone users globally will reach 3 Billion and mobile data traffic (cellular + WiFi) will exceed PC internet traffic for the first time.

The infrastructure used for 3G/4G is highly dependent on local access technology: i.e. a cellular network tower.

Vendors soon began to establish community-based WiFi networks to help throttle the load by installing lots of towers.

Throughout the years, various studies have been conducted to measure the energy consumption of WiFi-based smartphones. Despite advanced techniques, wireless traffic using 3G uses 15 times more energy than WiFi, while 4G consumes 23 times more. Thankfully, during the current COVID-19 pandemic, users are using WiFi-based networks. However, as already mentioned, this has caused a serious spike in residential electricity consumption.

Although desktop computers were (and are) usually connected to the internet via a wired link, laptops, tablets and smartphones are more often than not, wirelessly connected, either through WiFi or via a cellular network.

Given the newest cellular (5G) network currently under development, we now have the unique opportunity to identify research directions and tackle some of the energy consumption bottle-necks already at play. This is an incredibly important point and much debate has occurred already on the subject and implementation of 5G.

What about some potential solutions? Hmmm

Duplex Systems

Much of the challenges that lay ahead, such as building more efficient cellular networks, involve the design and production of more efficient air interfaces, alongside key transmission schemes.

Researchers are hoping to create a full “duplex” system that is meant to replicate a “self-sustaining wireless communication system”.

The advantage one has when incorporating a duplex system is a network that performs energy harvesting by itself.

This means having a base system identify groups of active users and splitting these groups down into smaller groups. Once this has been executed, the network begins shuffling groups across multiple-input-output channels concurrently.

Each user group receives data from their allocated respective channel transmit.

Groups of users are continuously shuffled and grouped via their current communication status, usage (time spent surfing) & geographical location. Internet providers have already begun implementing such methods but a majority remains a theory.

With more satisfactory analysis, the “duplex” system will inaugurate a new version of data consumption as we know it.

This approach may be effective when it comes to reducing data traffic and the overall energy intensity forced upon data centres around the world.

However, until internet service providers introduce “speed limits” to reduce user surfing habits and the intensity of graphical data exchange rates then active user energy consumption will continue to rise and enforce a greater strain upon our global digital footprint.

Dithering Models

Looking at this from a different point of view, another smaller-scaled option is to focus upon the content of the internet rather than the infrastructure.

One option would be to emphasise the importance of “Dithering” techniques - a digital compression method. Instead of using this for signal processing or audio wave analysis, as it is currently being used for - we would target graphical file formats such as JPEG or PNG and PDFs.

If successful, the next logical step would be to use this for audio & video content across every platform.

As discussed by Slavík and Prikyl., et al:

generally, all data contain some redundant information [...] The objective of a data compression algorithm is to transform redundant data into a form which is smaller leaving out the redundant information.

Data compression research is not new, as Slavík and Prikyl., et al highlight this when comparing the Dithering technique with other existing methods such as RLE Huffman coding and LZW (a traditional lossless compression technique most commonly used).

Another efficient technique is the Rice Algorithm, which is currently being adopted by researchers in compressing large “floating-point astronomical images”.

The Rice algorithm dithers the pixel values during the quantization process of compression. This “can greatly improve the precision of measurements in the images”.

Since these studies, researchers have released open-sourced software to help popularise these graphical compression techniques.

Internet browsers such as Firefox have incorporated image compression techniques but more emphasis on individual user education must be highlighted in order for us to profit from such research and make a significant difference.

New Global Policies

Considering video and audio are two of the largest contributors to our monster size internet digital footprint, we could outlaw the use of video and audio?

Why not apply equal “speed limits” for wireless networks? What about returning to a text and image-based medium of the internet?

What about allocating a specific energy budget toward the internet to help fund global “green” initiatives?

What about search engines displaying a numerical total page size value next to each returned search result found?

Such limits would not stop technological progress rather, it would ensure equal usage is monitored and scaled accordingly.

It would be a step in the right direction. Energy consumption throughout the digital age primarily rests upon our individual use of the internet.

Having universal policies in place to help reduce residential digital consumption rates wouldn’t necessarily be a bad idea. Such policies would include three important factors:

• Firstly, they would have a direct impact on the manufacturing, operational and disposal process of all devices that make up the internet infrastructure: end-user devices, data-centres, network and manufacturing.

• Secondly, they would have indirect effects on personal energy-use, due to the incorporation of “speed limits”. Such policies would impede a psychological effect upon users and would force them to think carefully about how they spend their time on the internet. More individuals would start to consider how and what the “internet of the future” would behave and look like. Such measures would reduce excessive audio and video consumption almost overnight.

• Finally, such policies would ensure that we as individuals “slow down”.

By this, I mean not just in reducing internet & digital energy consumption but more importantly, slowing down to regain balance within our lives and reduce our global dependency on digital life.

The term “slowing down” as used here, is merely a psychophysical structure or conception.

It is important to remember, that although the internet shifts consumption patterns, cater to technological and societal change, and contributes toward economic growth; our modern technological manipulation of the primary world, although unlimited and irresistible, has contributed towards overstrained modes of living.

By “slowing down” individuals will be able to fall back upon primal instincts, seek closure between loved ones or friends, and allow us to focus on better conditioning our general well-being.

Anyone well-read in history would quickly point out to us that humans have often made the mistake to extrapolate the future based on some “present development”.

By focusing all our energy upon this so-called technological “present development” as societies and individuals, this singularity leads us towards forgetting how to better focus upon the “present” within ourselves.