This month: Will battery charging halt the rise of electric vehicles? The Internet is actually addictive. DARPA wants to build DNA computers. Carriers are hyping 5G – does anyone (including themselves) know what they are talking about? And more.

Transport: Is charging the Achilles heel of battery vehicles?

What: By the mid 2020s, battery-powered electric vehicles are expected to have a range of the order of 500 to 800km and to be cheaper than today’s internal combustion engines cars, which will make them very attractive to consumers. If the world switched from petrol/diesel/gas to battery-powered vehicles, would the electricity grid be up to the job of charging their batteries? For example, there are 4.56m vehicles in Victoria [1] and a home charging station could draw up to 22kW of power, so if we assumed a load factor of 0.1, the total power demand in Victoria to charge electric vehicles would be about 10GW which is roughly equivalent to the current peak electricity demand for the State [2]. The good news is that this power demand would come when the normal load is at its minimum. However, what about parking garages where cars would want to be charged during the day? A large shopping centre has 10,000 parking spaces [3] and it if were to offer all them superchargers that were used simultaneously it would consume about 700MW which is roughly half the peak output of a large power station [4].

Implications: While the world will inevitably switch to electric drivetrain vehicles, whether those vehicles are powered by battery or fuel cell could have a significant impact on electricity supply. One thing is clear, however: electric vehicle charging will need to be controlled centrally which mean regulations, standards and technologies such as smart grid will need to be put in place before this gathers momentum.

[1] ABS

[2] Australian Energy Regulator

[3] Chadstone

[4] Energy Australia


Healthcare: The Internet is addictive

What: A team of researchers at Korea University studied 19 teenagers (average age 15) who had been diagnosed with ‘Internet addiction’, the symptoms of which are anxiety, depression and the use of the Internet to the point that it interferes with daily life. Using magnetic resonance spectroscopy of the brain they found that the participants had elevated levels of a neurotransmitter called gamma aminobutyric acid, or GABA, which has been linked with other addictions and psychiatric disorders. Twelve of the group were then treated with cognitive behaviour therapy, and after 9 weeks it was found that their GABA levels had declined to normal and that their use of screens had reduced [1].

Implications: The study only included 19 patients, so its results must be treated with caution, but it is an early indication that Internet addiction is similar to other addictions in that it is a real problem but is also treatable in the same way. However, this does not address the important question as to how prevalent is Internet addiction and who will be affected by it?

[1] IEEE Spectrum


Healthcare: Measuring patients’ vital signs without wires

What: Hospitals constantly monitor vital signs such as heartbeat, blood pressure and breathing of their patients. This is done with wired instruments attached to the patients, which inhibit their freedom of movement and sleep. A team at Cornell University has devised a method to measure these parameters without anything attached to the patient other than a couple of RFID tags which can be sewn into the patient’s pyjamas. The system uses a microwave transmitter to bounce signals off the patient. Movement in the patient’s body will modulate the amplitude and phase of the reflected signal. In theory this could be used to measure breathing but patient movement and other interference make this impractical. In addition, it would not be sensitive enough to measure heartbeats from inside the body accurately. The Cornell system thus adds 2 passive RFID tags, one on the chest and one the wrist to modulate and retransmit the signal into the body. The amplitude and phase of the returned signal can be measured to give an accurate reading of breathing, as well as the heart beat at the heart and the wrist. The time lag between these latter 2 signals can also be used to infer the blood pressure. The system can be used to simultaneously measure multiple patients from the same transmitter at the same time because each RFID tag modulates the return signal with a unique ID tag [1].

Implications: While this is a promising idea it is still just a research prototype. Even if commercialised it is unclear whether in a supplier-driven industry such as healthcare (designed around the convenience of the hospital and the doctors, not the patients) the benefits to customer will be sufficient motivation for hospitals to adopt it. Nevertheless, conceivably this system could be cheap enough for consumer adoption at home.

[1] Nature Electronics


Healthcare: Delivering drugs into the brain

What: The treatment of neurological disorders (depression, Parkinson’s and others) often involves drugs that change the levels of particular neurotransmitters in the brain. However, due to oral or intravenous delivery, the drugs are absorbed by the whole brain which results in side effects due to the levels of neurotransmitters in non-affected parts of the brain being changed. A team at MIT has developed a system to target the delivery of brain treatment drugs to the affected part of the brain only. The system consists of a microscopic tube (the width of a human hair) that is implanted into the skull and can go deep into the brain to the specific area being targeted. A sub-cutaneous miniature drug delivery pump is then able to send tiny, precise doses of drug down the tubes to the affected area.

Implications: This is still an experimental study and has only been tested on rats, but shows great promise in the treatment of brain disorders. Additionally, by placing an electrode on the tip of the tube, they can potentially do precise measurements on the impact of drug on neurological electrical transmission.

[1] MIT News


Mobile: Yet another 5G rollout claim

What: Carriers around the world are competing with each other to make 5G rollout claims. The latest is Spain’s Telefonica who say they will deploy 5G in Segovia (a beautiful city!) and Talavera de la Reina this year. Last week it was NTT Docomo in Japan saying that they would deploy 5G in 2020 [1].

Implications: The problem with this frenzy of marketing claims is that no one – least of all customers – has a clue what 5G is. The standards are not even complete. The carriers’ marketing departments trumpet Gbps speeds, 1-5ms latencies, and 100 times more connected devices on the network which are the headline goals of 5G, but the point is that not all 5G deployments will achieve these outcomes. None of the announcements indicates what speeds will be achieved on their actual rollout, which devices (mobiles, hotspots, etc) will work with the deployment, whether mobile as well as fixed services are supported, or the frequencies in which they will deploy the service (sub-6GHz or mm wave). Regarding the latter, while the high-speed Gbps claims are mainly based on the mm wave system, it is very difficult to support mobile devices at these frequencies because basically you need line of sight from the base station to the mobile device. The likelihood is that carriers are rushing to rebrand the evolution of 4G as 5G, just as they did with 3G to 4G.

[1] VentureBeat


Computing: DARPA is trying to build a DNA computer

What: It is well known that DNA has the ability to densely store vast quantities of information and retain the data for centuries. For some time, it has been possible to encode information in DNA molecules but, to date, there is no practical way of retrieving and processing that information. DARPA (the US Defence Advanced Research Projects Agency) is spending $15m to pay university research groups to find a way of overcoming this hurdle. The approach adopted by the University of Washington is to encode an image database of 10,000 images into DNA molecules. However, they are not encoding the pixels directly, but are first doing feature extraction on the images and then encoding the extracted parameters. Their aim is to perform image search on the database by encoding DNA molecules with the features of an image they are searching for (e.g. features of a cat). These molecules are then coated with magnetic nanoparticules and dropped into the DNA ‘soup’ which constitutes the database. The theory is that the query molecules will pair off with the DNA for the matching images, which can then be extracted using a magnet. Thereafter the relevant images are retrieved using a DNA sequencer [1].

Implications: While this approach has an almost science fiction flavour to it, success would open the door for a radical new form of computing which would overcome the power and miniaturisation limits that modern classical computers are hitting. However, it is likely to be decades before we see realistic DNA computers.

[1] Wired

Professor Hugh Bradlow is President of the Australian Academy of Technology and Engineering. He is recognised as a global leader in telecommunications technology, including being named by Global Telecom’s Business in the top 100 most powerful executives in the global telecoms Industry two years in a row, and by Smart Company as one of the 12 most influential people in Australian ICT.


to get daily updates on what's happening in the world of Australian Higher Education