Tech Trends 2024: Artificial Intelligence
Published on : Friday 05-01-2024
The present approaches towards artificial intelligence do not replicate any process or function of the human brain, says PV Sivaram.
Artificial Intelligence is a hot topic today, and people blessed with an abundance of natural intelligence are busy looking for some AI to use. It is a serious topic, there are big advantages being promised, which could be turned into big money for the early implementers. Having said that, where is AI being rolled out?
We encounter AI today largely as voice assistants on mobile phones or entertainment applications. It can respond to queries or prompts and attempts to largely emulate a human. But, for a tool which is so powerful, these are rather tame applications.
Some possible areas of investigation could be:
· Hype on AI today
· Subtle and steady ways of encroaching by AI
· Possible areas of application
· Present area of application
· High focus on development of algorithms and platforms
· Industry is not yet beating a path to the doorstep
· AI being seen as a threat to jobs
· Weaponisation of AI, and
· Fallibility of AI – possibility of AI making mistakes.
Intelligence and Artificial Intelligence can be studied by engineers and philosophers in different ways with a view to achieve different objectives. Here, we will examine AI from the point of view of Industrial Automation, with the manufacturing industry as the intended beneficiary. I will cover the topic in three instalments as below:
1. What is AI?
2. Where are we today?
3. Where shall we go tomorrow?
Image by Freepik
What is AI?
Why is it making waves?
AI is making waves! It is all over the news today, from tech journals to political news. We browse through this information, trying to get a grip of it all. Many times it doesn’t seem that the various people are all talking about the same thing at all! This article attempts to set a vocabulary of discourse, so that we at least understand each other before we launch off into a discussion on the more difficult topics.
AI is a very rapidly growing technology. It is an integral component of the Fourth Industrial Revolution creating a transformation of the manufacturing industry. However, the scope of AI stretches way beyond industry and manufacturing. It is subtly but strongly entering into every aspect of our existence.
AI has moved from a lab topic to a plaything for techies. Now it is entering into ‘serious’ topics. The days of AI playing games and AI enhancing capability of Tiktok kind of frivolity are now behind us. AI can seriously manipulate the transactions which we as individuals and organisations do with each other. In every discussion we need to be aware of the presence of one or more cyber-entities.
Debate raging today is whether AI is altogether benevolent? It does promise many benefits, but also brings in many threats. How to balance the negative with the positive? How to enable only the positive and block the negative? These are debates where every citizen should participate and express his or his opinion. But to participate and intelligently contribute, some knowledge is needed.
Is artificial intelligence a faithful servant? Or is it trying to become a tyrant, a master? Is it possible that the scientists in their labs are trying to create a powerful entity to serve mankind, which may escape and become a rogue?
There are some ways to visualise AI for humans. Since it is always software in a disembodied form, it cannot be provided as a box or gadget. The best way is to imagine a Cyber entity, which is always at the elbow of a physical (human or robot) entity offering decisions or suggestions relevant to the action or transaction that is going on.
We will restrict our discussion to examining the coexistence of Industrial Automation and Artificial Intelligence (IA and AI).
Terms and definitions
What is intelligence in the context of machines and Industrial Automation? We define Intelligence as the ability to take and retain instructions, retain the instructions, and to apply to different data as provided to the machine. This narrow definition of intelligence allows us to designate machines equipped with a stored program controller like PLC or CNC as intelligent.
Intelligence can be found in a device, in an algorithm, or a machine. The term machine includes robots. There is also intelligence obtained by clever arrangement of purely mechanical elements like gears and springs.
In the context of Artificial Intelligence, we need to broaden the view of intelligence. We can define intelligence in the context of machines as an ability to respond to situations which were not previously encountered, and emerge successful.
Programming of the intelligence – in earlier forms, it was embedded in complex wiring of relays. This principle of relays was carried over to PLC-programming, also in the form of relay-logic. As an evolution, this programming went over to High Level Languages.
Machines became intelligent by virtue of programming. A human programmer encoded instructions and deposited (stored) these instructions. The machine would interpret and execute a subset of these instructions in response to the stimuli received by it, and trigger certain actions, all of which were foreseen by the programmer. These machines are stored program controllers. Error-free operation was guaranteed only if the stimuli (or inputs) were among those foreseen by the programmer.
As the number of stimuli or inputs increases in quantity, it is not easy to foresee all possible combinations of occurrence of stimuli, and the sequence in which they appear. One way to handle this was to throw an ‘exception’ – equivalent to a machine saying I don’t know what to do, and raise an alarm or alert seeking human intervention.
Learning automata are such machines which can modify their own programs. The modifications arise from the feedback on the decisions made by them. This includes correction to the algorithm out of mistakes made, and reinforcement due to good results.
Machine Learning is a type of AI, where the Intelligence algorithm is not ‘programmed-in’. Rather the algorithm is derived from a labelled data-set also known as training data set. Here a large number of samples are provided with labels as ‘correct’ and ‘wrong’. The machine learns by studying the samples what makes a sample ‘correct’ and what makes it ‘wrong’.
Self-explaining AI is proposed that produces a decision and an explanation. It provides a human understandable explanation and confidence level for each decision and which was deemed to motivate and lead towards trustworthy AI. In general one cannot understand why the AI reaches the conclusion which it has come to. Therefore this optional feature is useful to make sense of counter-intuitive suggestions or decisions by AI.
Problem solving is a technique to identify a situation (a set of stimuli) which could lead to an undesirable state of the process, then determine a set of actions which could lead the trajectory away from the undesired state to a safe state.
The beginnings of AI
It was in the 1950s that a serious proposal was made to simulate the human brain by a machine. The first attempts were based on trying to emulate by creating devices to mimic neurons and interconnect the neurons to generate intelligent reactions to stimuli. This explains why AI literature is littered with references to neurons, including the very useful neural networks.
To be clear, the present approaches towards artificial intelligence do not replicate any process or function of the human brain. Early in the journey it was recognised that the human brain is far more complex than what was imagined, and the processes are yet shrouded in mystery. So this approach is kind of abandoned by engineers.
In the early days, machines which could repeat a set of tasks in the correct sequence without continuous human supervision were called automatic machines. Human intervention was required only to modify the programs and to repair and maintain the electrical and mechanical parts. This concept extends from single, self-standing machines up to a line of machines and entire plants.
What then is needed to convert an automatic machine to an intelligent machine? A capability for problem solving – that could be the answer. As the problems become more complex (more number of variables, less precisely defined safe/unsafe states), the number of responsive mechanisms should increase. The machine becomes more and more intelligent based on the number of responses at its disposal and the greater percentage of success in bringing the process into safe zone.
(This is Part1 of a 3-part series on Artificial Intelligence. Parts 2 &3 would appear in subsequent editions)
PV Sivaram, Evangelist for Digital Transformation and Industrial Automation, is mentor and member of steering committee at C4i4. He retired as the Non-Executive Chairman of B&R Industrial Automation and earlier the Managing Director. He is a past President of the Automation Industries Association (AIA). After his graduation in Electronics Engineering from IIT-Madras in 1976, Sivaram began his career at BARC. He shifted to Siemens Ltd and has considerable experience in Distributed Systems, SCADA, DCS, and microcontroller applications.
Sivaram believes strongly that digitalisation and adoption of the technology and practices of Industry4.0 is essential for MSME of India. He works to bring these concepts clearer to the people for whom it is important. He believes SAMARTH UDYOG is nearer to the needs of India, and we must strike our own path to Digital Transformation. Foremost task ahead is to prepare people for living in a digital world. He is convinced that the new technologies need to be explored and driven into shop floor applications by young people. We need a set of people to work as Digital Champions in every organisation.
For a deeper dive into the dynamic world of Industrial Automation and Robotic Process Automation (RPA), explore our comprehensive collection of articles and news covering cutting-edge technologies, robotics, PLC programming, SCADA systems, and the latest advancements in the Industrial Automation realm. Uncover valuable insights and stay abreast of industry trends by delving into the rest of our articles on Industrial Automation and RPA at www.industrialautomationindia.in
.jpg)
.jpg)
.jpg)
