Letters to the editor: volume 18, issue 7

Addressing rural ‘not spots’ should take priority over 6G

The article in the July 2023 issue of IET Member News regarding the ‘Why 6G?’ conference held earlier this year got me thinking. As a former chief engineer at Nokia Networks, I am 100 per cent supportive of research and development into mobile communication technologies. What concerns me is the growing gulf between R&D and deployment.

Where I have retired to in rural Wiltshire, there are vast areas where the best mobile communications technology available is 3G, not even 4G let alone 5G.

The planned switching-off of 3G – indeed, Vodafone has announced it will start switching off 3G from July this year – will leave most of the rural areas I drive through as mobile data ‘not spots’ with only 2G coverage. I’m not counting GPRS and EGPRS as credible mobile-data technologies in 2023.

The gulf between research into 6G and increasing areas of the country not having 3G, 4G or 5G coverage should be a cause for concern. Has mobile communications technology advanced beyond the ability of network operators to deploy these technologies, either through cost or NIMBY issues?

Should more research and development be directed into reducing the cost of these existing technologies so that their benefits are available to a wider population before we spend significant sums on technologies that are but a dream for those of us having to contend with the loss of 3G?

David Perry CEng MIET

Westbury, Wiltshire

Caveat EV Emptor

Two letters in the July 2023 edition of E&T describing EV owner experience piqued my interest. The first is from a Nissan Leaf driver, who highlights a range of systemic issues but overall loves the car. The second, from a Fiat 500e owner, discusses similar negative issues while feeling positive about cheaper motoring, especially with home charging and removing one less polluting car from the road.

In the case of the former, I do hope confirmation bias is not at play here, interpreting evidence in ways that are influenced by the beliefs that EVs are better for the planet. Many people, like the writer and myself included, who have been faced with fines for overstaying parking in a supermarket car park while trying to charge their EV, might well question the wisdom of transitioning from an internal combustion engine.

Filling the tank of an ICE vehicle is simple and quick. The ubiquitous petrol pump, filling at say 1 litre/second, or transferring approximately 10kWh of energy each second into the petrol tank, is essentially a 36MW power source. By comparison, one of the fastest-charging EVs, the Hyundai Ioniq6 Long Range 2WD, charging from an ultra-fast 350kW charger, has an average transfer rate from 20 per cent to 80 per cent charge of a mere 0.06kWh/second – a limit imposed by the battery, not the charger. Even Toyota’s recently announced solid-state battery technology, with a range of 700 miles and a charge time of 10 minutes, will only achieve a 5-6 times improvement over the present maximum energy transfer rate.

Why does this matter? Well, we all need energy in one form or another for our motoring needs, so EV battery energy-transfer rates compared with that of the common petrol pump give some idea of the number of ultra-fast charge points needed for hassle-free, every day, on-the-go motoring as we transition to an all-electric future.

The second letter mentions cheaper motoring, especially when charging from home. If everyone transitioned to EVs overnight, the Treasury would be short of around £40bn raised from the ICE motoring public each year. Surely this deficit will hit the EV motorist in some form and at some time with higher electricity prices. Beyond this, National Grid ESO estimates the cost of connecting an additional 50GW of offshore wind by 2030 at £28bn, while the UK government has in mind a figure of £140bn to achieve Net Zero by 2050.

These factors all suggest the price of electricity will only go one way in future. With this final thought and given our penchant for falling foul of the law of unintended consequences, we would do well to bear in mind the caveat emptor principle – let the buyer beware.

Dr Alan Jones CEng FIET

By email

The Cost of Charging Losses

I read with interest Toni Rivans’ letter on EV chargers, in particular the cost of public charging, in the July 2023 issue of E&T. In the UK, Ofgem policy is to meter the input to the charger so the EV owner pays not only for the energy to charge the EV battery but also the losses in the charger and connecting cable, which can be significant in a DC charger. This is different from other EU countries where the EV owner pays for the energy delivered to the vehicle, metered at the point of delivery (i.e. the connector socket). The EV charger owner is responsible for the losses and hence has an incentive to minimise them. Furthermore, unlike EU states, there is no requirement in the UK for the meters within charging stations to comply with the Measuring Instruments Regulations and there is no monitoring regime in place to ensure the metering is accurate.

CENELEC is working on a standard for EV charging systems to provide consistency across Europe and ensure that EV owners are charged accurately for the energy delivered to their vehicle. The UK is not involved in this working group, and with Ofgem’s policy of charging users for the energy losses, the UK EV owner will continue to overpay for public charging.

Mr Rivans is correct to state that public charging is a shambles. Failure of the UK government to regulate the sector means that in addition to paying for charging his Fiat he will be paying an unknown amount for the losses in the charger based on meter readings that may or may not be accurate.

John Cowburn CEng FIET

By email

Leave AI to Its Own Devices

The interview in the April issue of E&T with Dr Orit Halpern about her book ‘The Smartness Mandate’ sees smartness, via machine learning and artificial intelligence, evolving into a whole new way of life. This is inevitability already occupying many top management minds across the business world, coupled with underlying concerns about how to go about creating for their companies the most cost-effective and operationally trustworthy new lifestyle.

Such creative concerns, I have found in practice, are more likely to be laid to rest when business executives have at their fingertips a computerised model of the most appropriate technology-change path for the whole company to follow, rather than having to continually grapple with numerous departmental go-it-alone AI plans. Building such an organisation-wide model is a skilled, time-consuming, three-stage process. Initially, methodically unearth all critically essential business functions across the whole company. Secondly, rate each identified function from AI why, how, when, workplace relationship, cost/benefit perspectives. Thirdly, schedule all functions qualifying as serious AI candidates into an applications development and implementation project sequence most beneficial to company objectivity.

When each AI application is initially implemented, another concern often arises in the minds of line managers supervising nominated AI interface users. Will each user develop an effective continuing working interface with the AI algorithm, especially when checking and acting on its outputs? One way usually helping put such a concern to bed is to make sure user training is not treated as an implementation side-event: a well-known location for it on conventional computer application agendas. Instead, raise training to a level of importance allowing all AI interface users to become fully acquainted with their AI partner’s algorithmic functional twists and turns, including how to deal with any algorithm weaknesses inherited from coders that might show up in AI outputs.

Such human interface problems are completely avoidable when a smart thinking application is designed to work alone through automated direct links from and to its working environment. One shop-floor production-line application clearly demonstrated to me and others involved in its installation just how smart and trustworthy a smart-thinking algorithm operating alone in real-time can be.

George H Kelly MIET

By email

Heaviside’s Erratic Genius

I was very interested to read the ‘From the IET Archives’ article about Oliver Heaviside in the June 2023 issue of E&T. In about 1950 (the centenary of his birth), I read an IEE article which said Heaviside was a genius who, like Laplace, could only work at his best in a very warm room. At times while working at home, writing on any paper he could find including butcher’s meat-wrapping paper, if the gas meter stopped working his writing immediately changed to a tirade about the gas board and only resumed technically when another shilling was put in the meter.

AI McPhedran FIET

By email

The Lighter Side of AI

The perils of AI were foreseen many years ago in an edition of the 1960s US TV comedy show ‘Rowan and Martin’s Laugh-in’. Two scientists are examining a device. One says: “This is the latest product from our new automated factory.” The other asks: “What does it do?”. The first replies: “We don’t know, but when we plugged it in, it built a new factory.”

Tony Meacock MIET

By email

Write to E&T at engtechletters@theiet.org. We reserve the right to edit letters and to use submissions in any other format.