Enabled by massive advances in Natural Language Processing (NLP) and Natural Language Understanding (NLU), chatbot launches have exploded recently thanks to tech giants such as Facebook, Apple and Amazon that have opened up their messaging platforms to developers.
While the scale of development of these sophisticated, unmanned algorithms that can “talk” to users has been massive — three months after Facebook announced its platform for building bots, Facebook Messenger had more than 11,000 bots — not everyone thinks that chatbots are the answer to all our problems.
Ronny Fehling, who heads initiatives in artificial intelligence and cognitive computing at Airbus, believes that a chatbot needs to answer a couple of classic tests: First, is it better, cheaper or faster? If not, then intelligent search via a website or app is often a better option; Second, does the chatbot try to replace tasks that people love doing or hate doing?
“In the simplest terms, you could say chatbots are Artificial Intelligence-based applications that automate processes,” he says.
“At Airbus, chatbots are seen as a part of Robotic Process Automation (RPA); they should strive to automate processes in a business-to-human interaction context and should aim to replace those repetitive tasks that people hate doing.”
Fehling says that, for Airbus, the value in chatbots is their ability to use machine learning (ML) to deliver domain knowledge: “The chatbot needs to be trained on three things: 1. domain taxonomy; 2. data content i.e the information you actually want to search for, and 3. interaction models. The data content is typically ingested and represented by a knowledge graph, which has to be created first.”
To help identify where chatbots can help, Fehling has created an online use-case template that aims to help evaluate the business impact and usefulness of a particular potential chatbot use-case.
He is, however, cautious about the new technology and suggests that currently, most chatbots are not successful.
“We are definitely in a hype cycle,” he says, adding that largely chatbots have proved useful in a consumer setting where they can reduce the volume of queries that a support service receives.
He adds, “There are a couple of cases where chatbots really make sense; when you have long conversations, a lot of frustrating back and forth, typically over email, between humans. That’s really frustrating and time-consuming. Then, there are the interactions that are similar — in HR for instance where lots of people ask the same or similar questions.”
Simple frequently asked questions (FAQs) are a great place to look to develop chatbots, according to Fehling, who says that they could help Airbus to a certain extent with maintenance issues dealt with by the tech request centre.
He says, “A support engineer typically has to ask a lot of questions before he can provide an answer but he could have a fully fledged conversation that could take him to an answer that he can then verify. The system prepares a suggested answer and so, if it is correct, we reduce the cognitive workload for our support engineers and the time to closure for the support ticket.”
This, Fehling adds, would be really useful when a plane engineer needs some out-of-hours support from Airbus. While it may not provide a verified answer, a chatbot system can react immediately and help cover a lot of ground.
Head of R&T at Airbus Customer Services Laurent Pouget is currently working on a project to do just that. “We are trying to reduce and streamline the hours spent supporting our customers. We have portals and hotlines for customers to use when they have problems anywhere in the world and, of course, we have people behind these portals and hotlines who work to answer their questions and support our customers,” he explains.
“Supporting our customers is a duty and we always want to improve this activity. We spend a significant amount of money for that. In particular it is my intention is to use chatbots to keep improving our support to the customers.”
A chatbot in the cockpit
Airbus Research & Innovation Chief Jean Jacques Toumazet is also involved in a chatbot project, one that could lead to the introduction of a virtual assistant in the cockpit to help pilots resolve any of the complex situation management issues that arise in flight or on the ground.
Toumazet says that the solution will have to be speedy and succinct – chatterbots will not work: “From a pilot’s point of view, when they are on board, they need the answer to their questions in the smartest way as possible and as soon as possible. It is not ideal to have a long conversation. So the chatbot doesn’t need to be as developed in the conversational area as in other areas.”
He sees a virtual assistant as providing a speedy solution to pressing operational situations. He says, “For instance, perhaps you need to divert to another airport and you have 250 passengers on board. Does this airport have the correct hotel capacity? Maybe the best trade-off is to divert to another airport that can accommodate all the passengers.”
Toumazet also envisages a chatbot or virtual assistant that can anticipate problems that could arise on an airplane even if they are not actually happening at the time.
“For instance, we have studied hard landings, which is when the aircraft touches down at high speed and as a passenger, you feel it. We are looking at how to predict when the pilot might make a hard landing. This is not something you can predict with a classical analytical equation, it is really to do with observing the way the aircraft is approaching the runway and the way the pilot is behaving.
“We want to develop some machine learning that can warn the pilot that if they carry on the way they are they could have a hard landing. This is not really based on interaction like a chatbot but a system that is able to understand what is going on and then warn you without waiting for you to ask a question.”
A chatbot for every situation
On the ground, Michael Sillus, who works on innovations at Airbus in Hamburg, is working on developing a “chatbot for the shop floor” that will support workers by providing help with technical questions.
Part of Airbus’ vision of a “factory of the future”, he says, “We could have a bot that can help navigate through questions and help find the right answer to de-block a manufacturing situation, for example. A kind of first-line, design organisation support inside the factory.”
Sillus says Airbus has set up an online chatbot community to spread awareness of the technology: “It is important to have a point of focus to attract people to learn about the things we are doing because normally people get their information about chatbots outside of Airbus.”
He also points to a chatbot currently underway that will help project managers consider also sustainable development goals (SDG) as laid out by the UN.
“As a first Proof of Concept (PoC) we are building an “ecoDESIGN BOT” to help project managers integrate sustainability aspects “on the fly”. It will be integrated into a framework of other bots and help the project managers ask the right questions and provide the data needed to support the right decisions and thus avoid expensive and time-consuming Life Cycle Assessments (LCA) afterward. We are doing it with the innovation team here and a sub-contractor and with those who are experienced in the processes and environmental systems within Airbus.”
For Pouget, the long-term goal is to develop a system that can create chatbots across the company: “We are convinced that we are at the very beginning of a very important part of the company and of society. It is like a goldmine. It is something that we are convinced will change things very much, like when Microsoft delivered Windows or Apple started.
“We believe that the rise of chatbots will change how we work, communicate, access knowledge. Today, nearly everything is digitised but the problem is that it is a black hole; we can’t take advantage of it because there is too much information and it is difficult to handle it. The chatbots will be able to give value to what is digitised.”
Pouget believes that chatbots will play a greater role in the future at Airbus once the technology matures: “In the future, I see very few limits to the use of chatbots, which can be used for internal daily support for any Airbus employee. In our daily work we need to use information, we need to discuss things and we can be supported by a chatbot installed on our computer or smartphone. My ambition would be to equip Airbus employees with chatbots and progressively increase their scope of work for, say, human resources, for administrative information, for engineering knowledge management etc.”
He warns, however, that chatbot technology needs time to mature: “If we go too fast, they will fail and it will take years to come back. My plan is to make sure that the first chatbots are really efficient so that it is considered a success and thereafter development can go very fast.”
To achieve this, Airbus is looking to partner with outside specialists and build a culture of “chatbot” awareness within the company. Toumazet says that the best project teams are a blend of data analytics specialists and aerospace experts.
“Data analytics specialists are not necessarily knowledgeable about aircraft or flight dynamics but they know how to find things in the numbers. In the same team, we put people who know the way an aircraft behaves but don’t know anything about data analytics or statistics or machine learning. Brought together, these two different kinds of people will do wonders.”
“It is really difficult to take a data scientist and also turn them into an aerospace engineer and vice versa. What is interesting is to make people sensitive to the other discipline and this is exactly our way of working across the company and within our digital transformation.”
As a consequence, Toumazet says Airbus is looking for people with either a data science or aeronautical background, who can span the two disciplines: “Brought together it makes a perfect team with different skills aiming at the same goal.”
If that sounds exciting to you, why not work with us and be part of the teams that are building the future of flight?