If Moore’s Law holds true, artificial intelligence technology should be on an 18-month cycle of complete transformation. If recent investments, advancements and discoveries are any indicator, AI might have made Moore’s Law obsolete with the current pace of change.
In our last installment of Adventures in Artificial Intelligence, we revealed human service tasks that AI is poised to replace. These included everything from driving and dispatch, to customer communication and scheduling. Today, we turn our attention to the key opportunities AI will provide in improving personalization in field service.
Recent Artificial Intelligence Developments & Their Impact on Field Service Personalization
Google Program Beats Chinese “Go” Prodigy
As you likely already know, getting computer programs to react to changing real-time human behavior and stimulus is no easy task. While closed computer programs, like IBM’s Watson, have beat champion Jeopardy players as early as 2011, they didn’t truly think, process, react and learn based on information the same way humans do. Rather, they demonstrate the ability to store immense amounts of data, and surface the right piece of data at the right time.
But newer technologies and programs developed by the likes of DeepMind and OpenAI in recent months are beginning to crack the cognition nut. Google’s DeepMind team created a program dubbed AlphaGo designed to compete against 19-year-old Chinese prodigy Ke Jie, who is the top ranked player of the game Go; a board game widely considered to be the world’s most sophisticated.
So, is Go truly that much more complicated than Jeopardy, or even chess? Dutch computer scientist Victor Allis predicts that the average game lasts roughly 150 moves, with 250 choices for a character per move, which gives Go a game-tree complexity (used to measure the complexity of a game) of ten to the three-hundred-and-sixtieth-power. Which means, the possibilities are nearly endless.
Go requires a computer program to learn millions of potential moves, counter moves, and more importantly adjust strategy based on an opponent's actions in real time.
It’s one step closer to learning, a key hallmark of the human brain.
What AlphaGo Means for Field Service
While board games may seem irrelevant to field service management strategy, the software capabilities of AlphaGo are anything but. AlphaGo’s algorithm leverages deep neural networks and tree search technology. This software’s ability to predict, analyze, and act in real time are leagues beyond what was possible even two years ago. So how could programs like AlphaGo impact field service management in the coming years?
Hyper-Personalized Customer Prediction & Service
If AlphaGo’s ability to analyze data were unleashed on field service databases, customer service records, and even weather, customer location, or age; a level of automated personalization could be achieved that was previously thought impossible through computer programming.
This could include everything from learning exactly what service scenarios would incentivize customers to give your organization a positive review, predicting what color shirt your customers prefer that your field engineers wear, or even predicting what time of day they would prefer their service to be scheduled.
In all, the predictive modeling could prove immensely valuable in both improving service satisfaction and gaining faster customer resolution.
AI Chatbot Makes Personalized Skincare Recommendations
HelloAva, a new AI-driven chatbot that was recently launched promises to replace dermatologists’ recommendations for skin care products, with more environmentally friendly options.
Using text-based phone interactions or Facebook Messenger, users can sign up by answering 12 questions that are similar to what a dermatologist may ask patients during a visit. The data is processed, and the chatbot then categorizes the user into one of 30 different skin types, in order to make hyper-targeted skin care product recommendations.
So what exactly does a skin care chatbot have to do with field service management? We’ve certainly ridiculed chatbot fails on this blog before. But, this appears to be a more promising application of chatbot technology, with a sensible amount of machine learning and AI baked into the functionality.
Here’s what service organizations can learn from HelloAva:
- Customer Segmentation is Key to Unlocking Personalization
Satisfying field service customers has never been more complicated, with demographics and psychographic data ranging broadly across customers. In addition, services like Uber, and Amazon have drastically reshaped customer expectations. With this new territory comes broad challenges. Dozens of various customer groupings want completely different service levels from the same organizations. But often field service teams are only set up to provide service in one, or two ways.
Taking a cue from HelloAva, field service teams could develop chatbots that uncover deep customer preferences early in the service process, and use those to drive increasingly personalized customer interactions, and predictive maintenance.
- Applying Machine Learning Techniques Creates Deeper Personalization
Machine learning libraries give artificial intelligence algorithms access to pre-baked statistical models, and massive amounts of data. Artificial intelligence algorithms can use real-time customer inputs (like what is gained from the HelloAva questionnaire) combined with statistical models to output highly unique, and customer-centric service recommendations.
These recommendations would otherwise take data science teams weeks to manually develop.
The field service providers who tackle the machine learning challenge fastest stand to gain a much higher level of personalization among customers. While this might sound farfetched, in just a few short years it will likely be the norm.
AI & Autonomous Vehicle Market Spikes
A recent Research and Markets study reports that the autonomous vehicle market is forecast to reach $126.8 billion by 2027. These leaps in autonomous driving technology are powered by a mix of advancements in both road-based sensors and artificial intelligence (AI).
While companies like Uber, Volvo, Daimler, and even Google’s Waymo are all desperately trying to build self-driving trucks, the consumer vehicle market is surging ahead much faster.
Despite a great many field service publications citing a reduction in danger, or an increased bottom line as the biggest opportunities provided by autonomous vehicles, we believe greater customer personalization is the holy grail of a driverless fleet. But will field service organizations and technology providers stand by as tech giants pioneer driverless cars? Or will service organizations develop unique functionality and software that will set their service fleets apart?
Here are just a few ways an autonomous fleet could offer greater personalization for customers, and act as a strategic differentiator for service organizations:
- En-Route Customer Communication
If a field engineer were freed up from driving, they could effectively call, or even video chat with a customer while en route to a service location. This would allow the customer to describe their exact problem in more detail, or even display via video the problems they are facing with equipment. This would further familiarize the field engineer with on-site challenges ahead of the curve, and he or she would come prepared to hit the ground running.
- Self-service Vehicles
Just head to YouTube and search, “how to fix my dishwasher” and you’ll find thousands of results, both good and bad. This is evidence of a growing number of consumers who wish to perform routine service tasks on their own. This assumes those tasks are straightforward enough for the average consumer.
In the future, field service organizations could be sending autonomous service vehicles that contain basic equipment, and instructional materials to those consumers that wish to perform basic self-service on dishwashers, light fixtures, and more. While the majority of consumers prefer to have professionals perform this service, there’s a growing number of DIY customers rolling up their sleeves in today’s marketplace.
And why not lend them a hand?
- Autonomous Parts Vehicles & Drones
First-time fix rates are ideal for both the customer and the service team. Achieving first-time fixes means efficiency for the service team, and immediate resolution for the customer.
But even in today’s technology-driven landscape, it’s all-too-common for a technician or engineer to arrive at a customer location only to find that he or she doesn’t have the right part, piece of equipment, or device necessary for troubleshooting a problem. Whether drone-based, or road-based, autonomous vehicle technology opens up a world of possibility for field-based parts delivery.
Just imagine if a field engineer arrived at a customer’s location, didn’t have the right piece of equipment to finish the job, and could request the part be delivered within the hour via drone, or autonomous vehicle? While this may add an hour to the service visit, it might shave a week off the overall service resolution timeline.
Despite relatively little change in our daily service activities, artificial intelligence is making waves in dozens of areas. Whether it’s a game-based algorithm like AlphaGo, or autonomous driving software from the likes of Volvo, we stand to learn a great deal from AI advancements. But more importantly, we stand to improve customer experiences through a sensible approach to applying these new technologies.
To read more about artificial intelligence and all things field service, please subscribe to Field Service Matters blog.