Underlining a growing need for ‘digital literacy’

Digital transformation (DT) is the process of adoption and implementation of digital technology by an organisation to create new or modify existing products, services, and operations, by translating business processes into a digital format.1 DT is not an option, it’s an inevitability. Bearing this in mind, it’s in all our interests to be prepared and engaged in the process. The end-point we currently imagine is Artificial Intelligence (AI). This paper starts with the personal journey of Dr Nick Hill FIHEEM, director of Water Quality London, into the digital world, presents data based on a survey of healthcare estates professionals, and concludes with an opinion of where we should go from here.

At secondary school, my elder brother owned something called a ‘slide rule’ (for the younger readers of HEJ, you might have to Google that). I assumed that in due course I would also be the happy owner of such a device. Not so. In 1972, one of my classmates obtained a kit to build his own Sinclair calculator, at a cost of £56 (equivalent to £945 today). When assembled, it was capable of addition, subtraction, multiplication, and division. We were all amazed. Fairly quickly, prices plummeted, and we all acquired calculators. Leaving school, off I went to become an undergraduate biological scientist. However, I recall at that time that the only computer on campus was a mainframe in the Computer Science Department, into which would venture those students that most of us considered weird, clutching punched cards which apparently were fed into this machine. People muttered words like ‘Fortran’, but all this ‘computer’ stuff formed no part of my undergraduate degree; we had paper and calculators (albeit with statistical functions).

A year without interaction with computers

A year working for a water authority followed, during which time I had no interaction with computers. The next step was a Master’s Degree in Analytical Chemistry, still with no computers, but for some reason we went next door to the Royal College of Art, for a presentation on the ‘Information Superhighway’. Despite being surrounded by people with brains, this lecture went completely over our heads. We could not imagine the world being described. Rather than being excited or astounded, it was just too ‘far out’ to comprehend or believe. We couldn’t imagine it as having any relevance to our future lives. When we needed to do a literature search, we cycled to multiple libraries, not the internet. I then studied for a PhD in Public Health Engineering — during which time something that looked like a PC, but which was actually a word processor, arrived in our research section, although it was only operated by a secretary. Our ‘cut and paste’ was literally that; write on paper and edit it by cutting bits out with scissors, stick it to another sheet of paper, and — when finished — present it to the secretary for typing, the output being paper, not digital. However, when I popped out into the world of work again, I shadowed a Sales/Service engineer for a day, before being interviewed for a similar role. When I asked what he didn’t like about his job, he said ‘the admin’, which involved paperwork and dealing with information provided on large sheets of dot-matrix printed computer output. He said he got 18 months behind with his admin at one stage, and when I enquired about what he did about it, he said he ‘chucked it all away, and no-one noticed’. The data obviously didn’t have much value.

A career in consulting followed, which included serving healthcare organisations, with my first exposure to the varying extent of ‘computerisation’. Those were the days when saving digital files was not trusted, and everything was printed out and put in the filing cabinet just in case it was lost from the computer memory. I visited an Energy Manager in a large acute hospital, and since I’d never met him before, started with some small talk. “I noticed the old computers you’re throwing out, in the supermarket trolley outside your office door,” I said, to which he replied: “Those are our new ones.” To collect emails he had to walk three quarters of a mile across the site.

The world of mobile phones

By now we were in the world of mobile phones. I had an analogue phone with a car kit and a ‘booster’. It was not long before the mobile phone company kept calling me saying: “You can have a free new digital mobile phone.” To which I replied: “Great, and you’ll be fitting a new car kit for free as well, will you?” I would then be sold the benefit that my calls would now be private, to which I replied: “Who cares? I don’t mind who hears me say ‘I’m an hour from home, darling’, to which the reply is ‘I’ll put the dinner on.’ ” Eventually, we all succumbed to digital, and then smartphones. I read with dismay that children are arriving at their first school reception class unable to sit up because they have insufficient core strength, due to having stared at screens all their lives.²

Fast forward to the mid-1980s, and we had acid house music, which was accompanied by artwork based on fractals. At this time I became aware of a branch of science called the ‘maths of fractals’. For these not familiar with fractals, think of a fern, which has leaves on stems, which then divides into smaller leaves, and so on, in a spiral pattern. Nature is full of fractal structures. This was a time when my digital head really began to hurt. Some researchers took photos of classical statues in a formal garden, from a distance. They then enlarged the photos to reveal as much detail as possible, until they pixellated. They then took these digital photo files and applied the maths of fractals to them to predict more detail that might be present. When they re-visited the garden and examined the statues, sure enough all the predicted detail was present. This report coincided with a time when poor quality images of serious crime suspects were being sent to the FBI for digital enhancement, presumably using similar technology. Today we take for granted the manipulation of images by our smartphones.

Study for an MBA

My next sojourn into the academic world was to study for an MBA, as a mid-life crisis alternative to a sports car or motorbike. The course featured a module entitled ‘Strategic IT’. The lecturer contrasted Ford and Toyota. At the time Toyota was considered a far more successful company than Ford, and yet it spent half as much per car on IT. The American company had used IT for everything, but the Japanese had been selective, preferring to use, for example, coloured plastic discs where there was no perceived benefit of IT. The other interesting aspect of the lecturer’s research concerned the large government IT project failures, what caused them, and how they can be avoided. A key message was ‘Don’t do a massive rollout until a pilot-scale release has proved effective.’ It seems obvious really. Since then, the Government has embarked on a centralised patient record system that cost £10 bn (£3.6 bn over budget), which was the largest non-military system in the world. After seven years, only 13 out of 169 Trusts had received the full system, Barts lost thousands of patient records, and the Milton Keynes Trust wrote a cautionary letter about inefficacy and advised others not to use it. It was abandoned in 2013.³

The way data is generated and managed

I’ve previously expressed my thoughts on the way that data is generated and managed. Back in 2018, when ‘we’ve got a portal’ was all the rage, I was asked to present a paper on the subject of ‘Making the most of data for the benefit of public health’ at the Water Management Society conference.⁴ I harangued the audience as politely as I could on two of my pet hates — ‘abuse of spreadsheets’, and ‘portals as a haystack of PDFs’. The point of spreadsheets to me is to perform calculations, study trends, and plot graphs, etc but the data entry tends to be so poor that such operations are impossible without time-consuming data cleaning. My crude, but oft-quoted, phrase is ‘Don’t let anyone vomit into a spreadsheet.’ A more polite version is ‘rubbish in, rubbish out’, but in my experience it’s ‘rubbish in, nothing out’. Meanwhile, we have supplier portals where data are held as PDFs. Most of the time I find clients cannot even find their data, and of course cannot manipulate it. However, should something go wrong, as one ends up in court the supplier will be able to say ‘Oh yes, Your Honour, we provided those data in our portal, and gave our customer access’, thereby hanging you out to dry. Both of these examples concern a data science term, ‘dead data’ — defined as ‘data that is no longer useful’. It is commonly said that PDFs are ‘where data goes to die’. PDFs are designed to preserve visual layout, rather than the underlying data structure.

One of the most frustrating examples for me is laboratory testing data. For decades, laboratories have used LIMS (Laboratory Information Management Systems); basically large databases to track the samples they process, and the associated test results. More often than not, the output of the LIMS is a PDF test certificate, which is often provided — in the case of microbiological testing of water, to a water treatment company, which then transposes the data (with the risk of transcription errors) into a spreadsheet, in such a way that it cannot be analysed and manipulated. This is bizarre, when one considers that the LIMS is perfectly capable of outputting in spreadsheet format. So, instead of the data going from database to dead PDF, to dead spreadsheet, they could at least go from database to spreadsheet for trend analysis and more sophisticated manipulation.

An official opinion of ‘Where are we now?’ was provided in August 2021, when the National Audit Office published a report entitled Six reasons why digital transformation is still a problem for government.5 I’ll summarise them here, but expressed in terms of solutions, rather than problems:

• Understand aims, ambitions, risk.
• Engage with commercial partners.
• Develop a better approach to legacy systems and data.
• Use the right mix of capacity (people).
• Consider the delivery method (use agile methodology).
• Develop effective funding mechanisms.

The report highlights the need to understand the business problem before seeking a solution (i.e. no ‘trashcan strategy’). It emphasises the need to avoid unrealistic ambitions with no understanding of the associated risks.

In order to determine ‘Where are we now?’ at a local level, I’m guided by a paper written by Catherine Murray, in which she describes five stages of digital maturity.⁶

The 5 stages of digital maturity

1 Traditional

• Legacy systems, processes.
• Outdated ways of thinking.
• Little use of digital technologies.
• Lack the ability to drive change across the business.
• Activities that support DT are usually accidental, not a result of strategic intent.
• Likely being disrupted by competition.
• Must act quickly to build a strategic plan and organisation-wide awareness of why DT is critical.

2 Emerging

• Embrace digital slowly.
• Have modernised some aspects of their business.
• Largely reactive, and only make changes when they have to.
• Unable to outpace digital disruption.
• Must start addressing DT seriously to avoid creating more legacy issues.

3 Engaged

• Experiment with some critical elements of a winning DT strategy.
• Limited foundational activities and pockets of innovation are in place, but often siloed, and lacking focus or leadership.
• Need a plan for driving adoption of a singular digital vision.
• Key stakeholders must be engaged to develop a structured and sustainable transformation roadmap that delivers measured value.

4 Competitive

• Digital roadmap in place.
• Starting to combat disruption.
• Compete effectively in the current market.
• Need a strategy for future growth.
• Should start optimising and address any remaining blockers preventing them from launching and supporting new digital products or services that leapfrog competitors.

5 Maturing

• Have a well-established transformation roadmap in place that effectively fends off disruption, and evolves as needed.
• They use digital technologies to run their business, and have the ability to drive continuous change.
• Must develop a roadmap for continuous transformation and delivery, in order to realise their full potential and become leaders in their industry.
• Finding ways to remove friction enables them to react swiftly to market trends, and speed up delivery of new digital experiences.

Survey of Estates managers

In September 2023, I conducted a survey of 10 Estates Managers from different Trusts, using Slido — a software system which enables users ‘to conduct live polls to spark dialogue, check understanding, and get instant feedback’. The outcomes are shown in Figures 2 to 6. In terms of Murray’s digital maturity categories, the responses indicated that a large majority of Estates Managers felt that their departments were ’emerging’, i.e. slow, reactive, vulnerable to digital disruption, and must take DT more seriously. None of the respondents had a documented DT strategy in place, and some stated they would only have one if directed from the centre of the NHS. All of the participants said they used data for exception and assurance reporting, and a large majority for trend analysis, but very few used data for root cause analysis or reliability analysis. In terms of integration, none of those surveyed considered their IT applications, such as BIM, PPM, and BMS, to be fully integrated, although a majority indicated that their systems were partially integrated.

Three out of 10 said there was no integration. When it comes to the use of data, all of the respondents said that between 41 and 80% of their data collected is wasted, i.e. not analysed. This level of wastage would not be tolerated in any other aspect of the services provided by Estates Departments.

What about the future? I’ve said that DT is inevitable, but the secret is to get the most out of it, with the least pain. A couple of years ago I was in an airport looking at the books, not expecting to buy anything. Then Scary Smart7 by Mo Gawdat caught my eye. It was subtitled ‘The Future of Artificial Intelligence and How You Can Save Our World’. Mo is former Chief Business Officer at Google, so probably well qualified to comment. Bearing in mind that I was ‘off duty’, this was a bit of a heavyweight book choice, so for contrast I chose Expected Goals8 by Rory Smith, which is subtitled ‘The story of how data conquered football and changed the game forever’. I recommend them to anyone who wants to understand more about data and the future. The Scary Smart book spends the first half scaring the wits out of the reader, with three premises — AI is inevitable, AI will outsmart humans, and bad things will happen. The author observes that self-driving cars are already safer than cars driven by humans.

Self-driving cars

On a frosty day years ago, driving on the A40 dual carriageway towards London, I crossed a bridge, and the road curved to the right to become the slip road onto the M40. That’s when I spotted six cars all lined up alongside each other in the ditch, where they had lost control on the colder road surface of the bridge, and failed to make the bend. If those cars and my own had been self-driving, how many would be in the ditch? I think the answer would be one or none. As soon as the first car lost control, the data would be shared with the other cars (much as our satnavs share data) in order to allow them to avoid a similar fate. Alternatively, if each of the cars had data on the air temperature, surface temperature, accident history, and location of the bridge etc, the conditions would have been known or at least predictable, so as to avoid any of the cars ending up in the ditch. It’s easy to believe the data that demonstrate that self-drive cars are safer.

One of the most interesting statements was that comparing AI to human intelligence in 2049 will be like comparing Einstein to a moth. This raises some difficult questions, not least, ‘Would Einstein want to be subservient to a moth?’ The second half of the book explains how the potential nightmare scenarios described in the first half can be avoided. A further enlightening observation of Mo Gawdat is that the amount of data being generated is increasing rapidly, but that the volume of data analysis has stayed the same. This indicates a lack of systematic management. Correcting this anomaly should form part of your data transformation journey. Either use the data being generated, or don’t bother to collect it.

Trailblazers for digital technology

The Expected Goals book would certainly be relevant to anyone interested in data and/or football. One of the most surprising things about it is that the early adopters of data in football were ‘Big’ Sam Allardyce and Steve McClaren, unlikely candidates to many of us.

The moral of that story is that even if you perceive yourself as Sam or Steve (or even Samantha or Stephanie), it does not mean you could not be a trailblazer for DT. There have been plenty of examples of AI potentially going wrong. The old adage, ‘Anyone can make a mistake, but to make a huge mistake, you need a computer’, should be updated to ‘Anyone can make a mistake, but to destroy mankind, you need AI’.

There have been experiments involving computers working together, that have produced unexpected outcomes. In one case, the computers — while solving a problem — started to communicate with each other in a language they had invented themselves, and which the human developers could not understand, at which point the ‘off’ button was hit by the developers. In another example, the computers developed a new religion, at which point the experiment was stopped. It would be complacent to think that the ability to ‘switch off’ AI will persist. Would AI not be capable of anticipating the power being cut, would AI not be capable of avoiding that situation, or — if it occurred — taking measures to restore power? We live in an interconnected world, which AI would clearly recognise. Can AI not control robots and power plants? It is much easier to imagine that AI could self-perpetuate than the opposite. Perhaps a more amusing AI-fail I read recently concerns a children’s TV programme called Bluey. Bluey and Bingo don’t want to tidy their playroom, so they get Daddy Robot to do it for them, but when Bluey takes things too far and tells him that they never want to tidy the playroom again, he decides to throw the kids in the wheelie bin. Clearly Daddy Robot understands that the root cause of the playroom’s untidiness is Bluey and Bingo, and therefore takes the only logical step, which is to eliminate it.

There are a number of fundamental issues which will crop up during society’s move towards DT. One of them is ‘data democracy’. We are all familiar with the idea of ‘protected knowledge’. Our colleagues may try to keep information to themselves, believing it makes them indispensable, special, and an expert. Organisations act in a similar manner. It seems unlikely that in a world of AI, protected knowledge will survive. Our current belief that ‘knowledge is power’ is about to be undermined. The organisations that try to ‘own’ data may not survive or thrive. I often ask people how many personal surveillance devices they would tolerate. The immediate ‘without thinking’ answer is none, but within seconds the respondent remembers that data is being collected on them by their mobile phone, credit cards, car, TV, all the ‘Internet of Things’ devices in their home, the public transport system, ticket barriers, car parks, the buildings we enter etc, and of course myriad CCTV cameras.

Parting with considerable personal data

We are already parting with a colossal amount of personal data in return for the benefits of convenience, security, etc. However, this is one-way traffic mainly, and in the personal, not the commercial world. In future however, we should be providing and receiving data from suppliers, and sharing data more efficiently with peer organisations. This also means the end of ERIC returns. An immediate defence of commercial confidentiality arises, but ultimately when AI gets hold of these data it will be too late for a discussion about confidentiality. There will be resistance to data sharing for sure; the quote from The Cybermen is an unfortunate one, but ‘resistance is futile’. The GDPR⁹ and Data Protection Act¹⁰ have attempted to protect data privacy and security, and I’m sure were well intentioned, but I believe we will look back at them in much the same way that we laugh at the idea that a man carrying a red flag should precede us when we drive.

In 2023, I made a presentation to a group of Estates Managers, entitled ‘A Hitchhiker’s Guide to Digital Transformation’. The reference to ‘Hitchhiker’s Guide’ is homage to Douglas Adams’s Hitchhiker’s Guide to the Galaxy,¹¹ which featured Marvin the Paranoid Android, who complained that he had a brain the size of the planet, but was deliberately assigned menial tasks such as being responsible for opening and closing lift doors (think back to AI with the intelligence of Einstein, versus human intelligence equivalent to a moth). His book also featured the universe’s best supercomputer, which was asked the question ‘What is the meaning of life?’, and after years of deliberation, came back with the answer ’42’, which fits the idea that AI doesn’t always produce the answer one expects.

Quantum computer developed by Google

In 2019 Sycamore, a quantum computer developed by Google, solved a problem that the most advanced supercomputer would have taken 10,000 years to solve, in only 200 seconds. In June 2023 it was predicted that Microsoft’s quantum computer would overtake regular computers within two years.¹² The science fiction of Marvin will become reality, hopefully with less paranoia, and more appropriate deployment.

My survey (mentioned above) had a further three questions, under the heading ‘Where are we going?’ The outcomes are shown in figures 7 to 9. They reflect an optimistic assessment by the Estates Managers with respect to the digital future. The most popular response in terms of the consequences of digital transformation is that it will enable breakdowns to be predicted and pre-empted by maintenance activities. Currently, we rely on maintenance tasks performed at a frequency set out in manufacturers’ or other guidance. The data to support the selected frequencies may not exist, may be outdated, or may reflect a personal stake in service contract business. Artificial intelligence will consign this situation to history. Failure rates and asset status will be known, and failures pre-empted. Four out of ten respondents believed that DT will eliminate adverse patient outcomes related to the built environment, although around half believe that data collection will not be only collected by sensors, cameras, etc, but also by human beings.

Leadership of digital transformation strategy

In terms of leadership of their DT strategy, a majority of those surveyed believe that external consultants will lead, with only two out of 10 expecting internal leadership. When the benefits of a DT strategy were considered, the respondents continued to be positive, with a majority expressing the belief that there will better outcomes for healthcare, and less disruption of healthcare delivery. With AI, it will be possible to prove the impact that the built environment is having on patient outcomes. Those surveyed were most convinced that there would be genuinely risk-based PPM, continuous permanent improvement, and a solution to the backlog versus funding controversy. The Boards of healthcare organisations should take account of the groundswell of support — at least from Estates Departments — for the development and implementation of a DT strategy.

Where do we go from here?

I note that Prime Minister, Keir Starmer, tells us that the UK can double its productivity within five years using AI.13 It seems unlikely that an overnight AI solution to all our hopes or fears is going to be helicoptered in by NHS England or some other such body. However, the time is right for the adoption of a DT strategy to address specific business objectives. In order for the workforce to be ready, there is a need for digital literacy to significantly improve (where digital literacy is defined as ‘those capabilities that fit someone for living, learning, working, participating, and thriving in, a digital society’). There are some really simple shifts in approach that will smooth the transition to a data-driven world. What data to we need to collect? How are we going to analyse them? How are we going to ensure that data are clean and in a non-dead format? How can we use procurement to specify the data and their format when buying all goods and services? What data are we willing to share, and what data would we like others to share with us? If we can answer questions such as these, we can avoid a piecemeal, chaotic journey to DT. How many of your disappointments (e.g. contractor/product performance, poor procurement, building handover) could have been avoided if we had powerful data management?

Footnote

• The IHEEM SEMAP (Strategic Estates Management Advisory Platform) is forming a group concerned with Digital Transformation. I have been asked to chair the group, which is in the process of seeking members and determining its objectives. I would welcome your thoughts via email to office@iheem.org.uk

References

1 Wikipedia Contributors. Digital transformation [Internet]. Wikipedia. Wikimedia Foundation; 2019. Available from: https://en.wikipedia.org/wiki/Digital_transformation

2 Fewer than half of parents think Reception pupils should know how to use books [Internet]. The Mail. 30 January 2025 [cited 2025 Feb 2]. PA News. https://tinyurl.com/3cpm6ryh

3 Syal R. Abandoned NHS IT system has cost £10 bn so far [Internet]. The Guardian. 18 September 2013. https://tinyurl.com/yc6b85dm

4 Hill N. ‘Making the most of data for the benefit of public health’, presented at ‘Back to Basics and Forward to the Future’ conference, Water Management Society, 2018.

5 Gallagher Y. Six reasons why digital transformation is still a problem for government — National Audit Office (NAO) insight [Internet]. National Audit Office (NAO). 4 August 2021. https://tinyurl.com/45mj49rb

6 Murray C. The five stages of digital maturity: How does your organisation rank? [Internet]. businesschief.com. 19 May 2020. https://tinyurl.com/8am4phrr

7 Gawdat M. Scary Smart. Pan Macmillan, 8 December 2022.

8 Smith R. Expected Goals: The story of how data conquered football and changed the game forever. HarperCollins UK, 2022.

9 GOV.UK. Regulation (EU) 2016/679 of the European Parliament and of the Council of 27 April 2016 on the Protection of Natural Persons with Regard to the Processing of Personal Data and on the Free Movement of Such data, and Repealing Directive 95/46/EC (General Data Protection Regulation) (Text with EEA relevance) [Internet]. Legislation.gov.uk. 2016. https://www.legislation.gov.uk/eur/2016/679/contents

10 Data Protection Act 2018 [Internet]. Legislation.gov.uk. 2018. https://tinyurl.com/2kd9yrj6

11 Adams D. The Hitchhiker’s Guide to the Galaxy. London: Arthur Barker; 1979.

12 Cuthbertson A. Quantum computers to overtake regular computers ‘within two years’ after breakthrough [Internet]. The Independent. 22 June 2023. https://tinyurl.com/yjxbju6k

13 Cecil N. Keir Starmer gives full backing to Rachel Reeves as he claims AI can double UK productivity in a few years [Internet]. The Standard. 13 January 2025. https://tinyurl.com/5n9ajekk