“Digital transformation” has been around for a couple of decades and its popularity shows no sign of slowing down. During the Covid pandemic enterprises large and small have determined that they need to do more digitally in order to survive and so the trend towards ‘going digital’ has accelerated. The realm of digital transformation is so vast it covers anything from “moving to the cloud” to “we need a new, automated process” to “closing down physical call centers”.
Digital transformation can be explained simply as “strategically adopting digital technologies in order to simplify business processes, improving productivity, and efficiency, and drive new value for customers”. There is no one-size-fits-all digital transformation process that applies to all companies. Digital transformation also is not a quick fix project, but rather a longer-term process of change.
Digital transformation also has its own jargon. As more and more companies turn to digital, and the jargon becomes commonplace, this short article may clear up some questions or refresh your knowledge of the most standard terminology.
Additive Manufacturing
Additive manufacturing is more commonly known as ‘3D printing’ and has been around for 40 years but its popularity and accessibility has only skyrocketed in recent years. Additive manufacturing creates complex designs in less time and uses fewer materials as compared with traditional manufacturing or prototyping methods. A 3D printer “prints” by pushing molten plastic through a nozzle, layer by layer according to a programmed design.
Although commonly used for rapid prototyping and manufacturing of physical objects in sectors such as industrial manufacturing, dentistry, prosthesis creation, replications, pathology, movie props, and aerospace, additive manufacturing has also been helpful during the current global pandemic.
A hospital in Italy, for example, urgently needed respirator valves for COVID-19 patients but was unable to obtain supplies due to the country’s lockdown. An engineering firm worked with a local manufacturer to produce the required respirator valves within 24 hours using additive manufacturing and were able to save the lives of many Italian patients.
Another example from Canada: a 13-year-old boy scout used additive manufacturing to print and donate ear guards for frontline medical staff to help protect their ears from abrasion caused by the security strap of face masks.
Agile Methodology
The agile philosophy is an umbrella term that describes a results-focused approach project execution centered around adaptive planning, self-organisation, and short delivery times. Two of the most popular agile approaches are Scrum and Kanban. One of the key differences between an agile and a traditional approach is in its delivery. The traditional delivers solutions in the later stages of the project. The agile however delivers solutions in an incremental, iterative way throughout the process allowing for course corrections.
Artificial Intelligence
Artificial intelligence (AI) has been around for more than half a century. A term first coined in 1956 by John McCarthy, an American computer scientist, to define “the science and engineering of making intelligent machines, especially intelligent computer programs”.
The authors of A Modern Approach, Stuart Russell and Peter Norvig, explored four different approaches that have historically described the field of AI: thinking humanly, thinking rationally, acting humanly, and acting rationally. One of the classic AI examples is IBM’s Deep Blue, the computer that defeated the former world chess champion Garry Kasparov in 1997.
Narrow AI v. General AI
Two main types of AI are narrow AI and general AI. Narrow AI encompasses all types of Artificial Intelligence that we see today, including self-driving cars and digital assistants like Apple’s Siri or Amazon’s Alexa. Narrow AI can only execute the tasks they are designed to perform. It is pre-defined, specific and limited without the capability to be self-aware, conscious, or think for itself.
General AI is an AI agent which has the competency to think, learn, comprehend, and behave like a human. Google’s Director of Engineering Ray Kurzeil has predicted that by 2045 (only 25 years from now!) AI-machines will be more powerful and capable than humans making the notion that General AI applied as widely as Narrow AI in the relatively near future quite formidable.
Let’s explore some subsets of AI: voice recognition, facial recognition, natural language processing, machine learning, and deep learning.
Voice Recognition
Soon security questions, passwords, and one-time passwords (OTP) will no longer be secure enough to authenticate a person’s identity. Financial institutions like Barclays, DBS and Citi are already raising the bar by turning to biometric voice authentication to help their clients protect access to their accounts.
Facial Recognition
The Facebook conglomerate is actively exploring facial recognition using a curiosity-arousing machine learning detector. They call it ‘Deep Face’ and have trained the detector using four million images uploaded by Facebook users. Users may get notifications when someone adds a photo that might include them. This prompt to prevent identity misuse is one of the ways Facebook uses facial recognition.
Another example of facial recognition with which many Apple users may be familiar is the ability to unlock your phone via facial recognition. According to Apple, the possibility of unlocking one’s phone with a random face is one in one million.
The Government of the United States has also started using facial recognition scanners in certain American airports to improve the efficiency of identity verification. This scanning technology has a 99% matching success rate and takes only two seconds.
Natural Language Processing
Natural Language Processing (NLP) is a popular subset of AI. It is the ability of a machine to process and comprehend written or spoken content produced by humans. Chatbots, smart home assistants like Amazon’s Alexa and OK Google, as well as email filters that identify spam messages are made possible with NLP.
Microsoft Office 365 has a built-in editing function that provides spell check and grammar suggestions to improve the quality of text written by the user. In 2019, Microsoft announced an upgraded version of Editor – Ideas in Word, which not only serves as a proof-reader, but also includes justifications and explanations for each suggestion. New features include an approximate reading time instead of simply a word count, key points, etc.
Many companies also use NLP to screen resumes accelerating their recruiting process. It is common to parse resumes in various formats such as .pdf, .txt, .doc and .docx seeking predefined keywords according to different job postings.
Machine Learning
Machine learning is computer learning from past experience and therefore improved functionality with each iteration of a task. The machine learns through data and information but is prone to human bias, as the data input may not be neutral. For example, machine learning is used by police departments to predict and prevent crime, however there have been instances of false alarms and arrests due to biases of the person inputting data.
In the healthcare industry, machine learning is used to detect respiratory issues, cancerous tumors in mammograms and skin cancer. Uber, a ride-hailing company, uses machine learning to adjust ride price according to predicted rider demand and car availability.
Although machine learning takes less time to train, the results provided are less accurate when compared with deep learning. The data needed for machine learning must be structured, whereas deep learning can process and learn from unstructured data.
Deep Learning
Deep learning is a subset of machine learning that was developed to understand irregular data for tasks such as object detection, speech recognition, and language translation. Deep learning uses massive amounts of data generated by the general public and similar to machine learning, learns from past experiences and to improve but without predetermined instructions by programmers. Autonomous driving cars for example ‘learn’ from deep learning machines to recognise road signs and traffic symbols.
Big Data
Methods to access, store, and process large amounts of data have been around for a long time, but the term “big data” became popular in the early 2000s. “Big data” has to meet three criteria: volume – there must be a huge amount of information; velocity – the input, processing, storage, and retrieval of data must be rapid; variety – the data is not limited to a specific type or form, and comes from a variety of sources.
The ultimate value of big data is not how much you have, but rather how you use it. When big data is skillfully analysed, it offers the opportunity to increase sales and returns, can enable innovation and cost reduction, and improve efficiency.
Netflix & Amazon
Netflix, the global video-streaming provider founded in 1997, has 182 million subscribers worldwide as of April, 2020 and leverages big data to personalise recommendations to subscribers based on prior viewing.
Similarly Amazon, the e-Commerce giant, has a personalised recommendation system that depends on data collected about each user which is then analyzed and then to offer product recommendations about “buyers like you” which combats a buyer’s frustration when offered too many choices.
Cloud Services
Cloud Services provide storage for your company’s data, but do not require you to purchase and maintain servers, devices, or machines. The cloud is like warehouse space you rent, and the cost increases or decreases depending on the amount of space you use. Traditionally companies would store their data on servers they purchase, build, and manage – they would own the warehouse. In addition to the high cost of ownership is the risk one takes should traffic or space needs abruptly increase. Not only does the cloud mitigate the risk of ownership, but provides the additional benefit of simple disaster recovery and data access from any device via the internet.
Well-known cloud services include: Google Drive, Microsoft 365, Dropbox and iCloud.
Three Kinds of Cloud Services
Infrastructure as a Service (IaaS)
Some common uses of IaaS include web applications, website hosting, testing and development, and big data analysis. IaaS allows virtualisation, is highly scalable and users pay only for what they use. Some examples are Amazon Web Services (AWS), Cisco Metapod and Microsoft Azure. IaaS is the only cloud service that requires a tech-savvy person to control their infrastructure as it is a “self-service” option.
PaaS (Platform as a Service)
A typical user will seldom need to use this form of cloud service as it is usually used for applications or software developers. PaaS allows users to design and test out new applications without worrying about mundane tasks such as hardware maintenance, software updates, or security issues, and rather enables them to focus on creating and experimenting with an application. An example of PaaS is the Google App Engine.
SaaS (Software as a Service)
SaaS has perhaps the widest usage amongst all three forms of cloud services. SaaS usually comes with a monthly subscription, if not free of charge. SaaS provides security, maintains hardware, and does not require the user to install or run software applications as most SaaS applications are available through a web browser with an internet connection. One can log on to an account from any device as information is stored in the cloud. Some popular examples are Gmail, Yahoo Mail, iCloud, and Google Drive.
Customer Experience
Customer experience (CX) is not limited to customer service but to every touchpoint a customer has with your company. Offering a seamless and personalised customer experience is a key motivator for digital transformation. Conventional lines between customer interactions online (such as websites) and offline (such as physical stores) are quickly blurring and integration between the two are being made seamless by the use of technology, human touch, and data-driven insights. Customers are expecting more from companies – better service, faster online purchase processes – and CX is a key differentiator between businesses. Some examples of a poor customer experience include unstructured customer service, the lack of human touch throughout the journey, long wait times, multiple touch points needed to resolve an issue, as well as unhelpful employees. Improving the customer experience can take on a variety of approaches, from streamlining the customer journey for efficiency and enjoyability, to solutions as simple as offering multiple payment options, and providing improved employee training.
Disruption
Disruption is one of the widely used digital transformation buzzwords. Disruption is said to happen when a new company uses a new technology and positions itself as a threat to incumbent companies within a particular industry. Disruptive innovations do not happen overnight however by the time incumbents recognise and understand the challenge posed, it may be too late for them to develop a differentiated response significant enough to protect their market share.
Netflix
A typical story of disruptive innovation is the rise of Netflix and the fall of Blockbuster. When Netflix started in 1997 as a DVD (digital versatile disc) rental service, people were still renting VHS – home video cassettes. Blockbuster, which had 1,000 physical stores in the United States during the early 1990s, was thriving. Netflix services attracted a very small group of consumers consisting mostly of early adopters of DVDs and online shopping and the company approached Blockbuster with a buyout proposal, and were rejected. Netflix struggled to stay afloat, becoming slightly more popular when the DVD trend swept across the public. Changing course to take advantage of in-home internet use, Netflix is now the largest subscription streaming service globally, with a valuation of $196 billion (*as of April 2020). Blockbuster filed for bankruptcy ten years ago.
Wikipedia
Another famous example of disruption is Wikipedia, the world’s largest, free online encyclopedia that eclipsed physical encyclopedias. Wikipedia was founded in 2000 by Jimmy Wales who had a vision to make knowledge available to everyone whenever, wherever, without charge. The online encyclopedia is extremely convenient, unlike the physical volumes of encyclopedias, and is accessible by anyone who has an internet connection.
Internet of Things
The Internet of Things (IoT) is a term coined by Kevin Ashton back in 1999, but like many other technological concepts, it took some time for the public to understand the idea. Any physical object that can be controlled through the internet can be considered an IoT device. From a smart lightbulb to a smart city, IoT connects the physical and digital worlds. IoT is a fast-growing segment with seemingly endless opportunities.
Greentech is a sector that uses IoT extensively. Although these environmental technologies have not been around for a long time, they promote sustainable and clean energy production with a low carbon footprint. These environmentally friendly solutions, made possible with IoT, are common both in personal and industrial usage.
Personal IoT devices include wearable medical health monitors that store data online and then may be referenced by medical professionals to make more informed decisions about diagnoses or treatments. Personal IoT devices can also include smart home applications such as lightbulbs, thermostats, door locks, refrigerators, or home assistants like Amazon’s Alexa and Echo, Google’s Home Assistant, or Apple’s Siri. However, as IoT devices collect a huge amount of sensitive and personal data, one of the biggest concerns with IoT is security.
Smart cities apply IoT with traffic management decreasing traffic congestion. In retail, IoT is used as a smart inventory management system, smart customer tracking and to enhance the customer journey.
Agriculture and farming also integrate the usage of IoT into their daily life.
Cattle monitoring and management applications can automatically keep track of the animals’ health, activity, and temperature. These applications will also notify the farmers when the livestock roam from the herd, keeping an eye on the animals without the farmers wasting manual hours and energy.
A smart weather monitoring system is also widely used by farmers as it collects data about rainfall, temperature, humidity, wind and air pressure enabling better decisions on water resources, lighting usage, maintenance, and greenhouse simulation among others, allowing growth in a highly efficient manner.
Robotic Process Automation
Robotic Process Automation (RPA) is entry-level software for automation platforms which provides an alternative to manual, repetitive labour such as data entry and customer queries. RPA software has neither the capability to make decisions, but only to do what it has been programmed to do. RPA has been around for about two decades and is one of the technologies that has interested VCs recently.
Intelligent Process Automation
Intelligent Process Automation (IPA) is the combination of Robotic Process Automation (RPA) and Machine Learning (ML). Like RPA, IPA is used to substitute humans in doing repetitive work however, IPA goes beyond and can complete more complex tasks. It can also process a large amount of data in a short time, increasing efficiency and accuracy.