With Forrester predicting artificial intelligence (AI) investments will jump 300% in a single year, it’s obvious why AI is currently one of the hottest topics in business. Accenture sees big things on the horizon too, with an annual predicted 95% improvement rate in AI speech recognition technology.
There’s certainly no doubting the amazing impact AI advancements will have on businesses globally. When combined, AI, automation, big data, machine learning, and the Internet of Things (IoT) will revolutionize business processes, and consumer technology.
Will we have chips in our brains that can read our thoughts? Hire robot overlords to watch our kids while we’re on vacation? Hand our jobs over to smarter, faster, more perfect humanoid-machine hybrids? Don’t bite on that stuff, it’s just a bunch of hype.
In reality, big strides were made in the past year across chatbot technology, voice recognition software, medical applications, driverless cars, ecommerce technology, customer experience and more.
But what scholars, pundits, and even technologists can’t seem to agree on is exactly how AI will transform the everyday operations of specific businesses. This is frustrating for business leaders seeking straight dirt on how they can leverage this most hyped of technologies.
We believe it’s time for a realistic snapshot of AI, and how it can improve field service. Amidst all this noise, it can be helpful to step back and examine specific scenarios. AI should be embraced in the workplace, not feared.
In our new series, Adventures in Artificial Intelligence, we aim to dumb down this smarty-pants technology. In this first post, we offer a brief definition of artificial intelligence, and uncover the four key types of AI capabilities, and how they can be applied in field service.
Naturally, AI will take years—maybe even decades—to fully develop. In the meantime, finding small and profitable ways to apply new capabilities will prove the best use of our time and energy in service.
Artificial Intelligence Defined:
“AI is intelligence exhibited by machines. In computer science, the field of AI research defines itself as the study of "intelligent agents": any device that perceives its environment and takes actions that maximize its chance of success at some goal. Colloquially, the term "artificial intelligence" is applied when a machine mimics "cognitive" functions that humans associate with other human minds, such as "learning" and "problem solving" (known as Machine Learning).”
But, this definition still leaves much to the imagination.
Let’s turn to Michigan State University scholar Arend Hintze, who recently broke AI down into specific categories in his article featured on Government Technology Magazine’s site.
Below are four types of artificial intelligence technology, as defined by Arend, and ways each can be applied in field service.
4 Types of Artificial Intelligence
1. Reactive AI
Reactive intelligence software can perceive the world directly, and make programmed decisions based on observations. This type of software can not form memories, and does not have any real perception of reality. The most well-known example of this is Deep Blue, IBM’s supercomputer that beat multiple chess grandmasters in the 1990s. As this technology evolved, it was featured more recently on Jeopardy, where it also beat out top players.
In field service, reactive AI can be used to monitor parts inventory and inform logistics professionals on how to improve efficiency. It can also be used to automatically deliver customer details to field technicians via text message, email, or even via voice technology. Some service organizations are leveraging this type of technology to automate aspects of service scheduling and dispatch.
But in the end, reactive AI is ultimately programmed. It won’t learn, adapt, or take over your brain.
2. Limited Memory AI
This type of AI technology observes, logs, and reacts to short-term insights gleaned from real-life situations. The most common applications can be seen in driverless car software which logs speed, lanes, weather, traffic lights, and other factors that require short term memory. For example, observing lane markers and road speed for several seconds can yield enough data for AI to control a safe lane change, without any driver assistance.
But limited memory AI ends at that. These systems can’t remember data points as long-term experiences to be used as reference points in similar situations. As in, they don’t actually think. This is where current AI capabilities stop, and human brains begin.
In field service, limited memory AI could have obvious applications, one of which was already mentioned above: a driverless fleet of vehicles.
A second useful application in field service will be combining augmented reality glasses, and limited memory artificial intelligence capabilities. In the same way driverless car technology uses sensors and cameras to know where lanes and roads go through, augmented reality glasses could be programmed to see equipment parts through visual, temperature, or even thermal perspectives.
Smart glasses could be worn by techs, and used to better direct them towards the swiftest equipment resolution through the viewing, scanning, and measuring of field-based equipment. Smart glasses could identify issues that would have taken the tech significantly more time and effort to resolve.
If artificial intelligence can identify and help resolve service issues in moments, instead of minutes, field service could be transformed forever.
3. Theory of Mind AI
Theory of mind AI takes us deep into future territory. These types of artificial intelligence software and machines would be capable of forming technical perceptions about their surroundings. They would also apply long-term memory learning, and be able to adjust their behavior based on external stimuli. The external stimulus could be anything from a facial expression, to the local air humidity. In essence, machines would be able to understand how people, creatures, and objects in the world impact their own behavior. In psychology circles, theory of mind is described as a big stepping stone towards self-awareness, self-control, and autonomous decision-making in humans.
Down the line, this type of AI could be used to completely automate customer interactions. Theory of mind AI service software could call a customer, ask them a series of questions, and be capable of adjusting their tone of voice by interpreting the tension, or joy they hear in a customer’s voice. In addition, they could pull from a database to make service recommendations based on past customer problems, and preferences.
4. Self-aware AI
Beware, we’re now entering into a Hollywood fantasy. Films like Star Wars, Transformers, and Ex Machina all show us fantastical versions of self-aware, human-like robots. Whether it’s the huge machines protecting the universe, or evil robots taking over every galaxy, this vision of artificial intelligence most certainly does not exist.
In theory, future self-aware AI systems and technology could form ideas, thoughts, and representations about themselves. Instead of being programmed to mimic cognitive functions based on external stimuli, they would quite literally think.
Here’s how it’s different.
The theory of mind AI might enable the following process in a robot: “That furnace is broken, so I should fix that furnace.”
But self-aware AI would enable the following: “I know the furnace is broken, so I should fix that furnace, and find out why it is broken.”
Machine becomes man when they ask, “why?”
But will that ever actually happen? Tough to say.
Check back in for next week’s Adventures in Artificial Intelligence where we’ll turn our attention to virtual assistance technology. When paired with AI, assistance tools will be some of the first to significantly improve technician efficiency in field service. In that piece, you’ll learn all about virtual assistance opportunities, and how you can prepare to scale this new technology and streamline service.
Keep an eye out for future Adventures in Artificial Intelligence posts by subscribing to Field Service Matters.