AFRY develops chatbot for Sahlgrenska University Hospital

Later this fall, the biggest hospital in Sweden, Sahlgrenska University Hospital in Gothenburg, will test a conversational AI chatbot solution to provide support to their patients. The project is funded by VGR Innovation Platform and Vinnova. AFRY has developed the chatbot, which will be deployed for testing in one of the nursing wards at the hospital as the first step.

Chatbots can be thought of as the spokesperson for artificial intelligence (AI) which can interact with anyone in any language. It works as a computer program that attempts to simulate the conversation or “chatter” of a human being through text or voice interactions. For example, a user can ask a chatbot a question or make a command, and the chatbot responds or performs the requested action through its application interface. A question to the chatbot could be something like “When will my visitors arrive?” and the chatbot would answer “Visiting hours are between 5pm and 7pm”.

For a hospital, having a chatbot is very beneficial due to its time and money saving qualities. It can respond to the patient’s queries in a fraction of seconds while the human nurse’s time can be allocated on other tasks.

– In such as big hospital as ours, providing this service is really important. We have seen a growing demand from both employees and patients to simplify their correspondence and with this test we are trying to develop a chatbot meeting these need, says Yvonne Hedegärd, Strategist and Project Leader at Sahlgrenska.

From food to personal health

When the chatbot has been tested and implemented, it is supposed to be available around-the-clock and provide answers and solutions to the patients’ queries, which can be ranging from everything from food or location to personal health. However, if a question is marked ‘private’, it will be handled by a human nurse.

– What’s really great about this project is that we use the latest AI technology to create real value for both Sahlgrenska and the patients. The first tests have been successful, and now it has to hit the road by being deployed in real settings. That is how AI projects work, says Prajit Datta, AI Research Scientist at AFRY.

Natural language processing

At the heart of the chatbot technology lies natural language processing (NLP). The chatbots process the text presented to them by the user before responding according to a complex series of algorithms that interprets and identifies what the user said, infers what they mean or want, and determine a series of appropriate responses based on this information.

Some more facts

• Chatbots are becoming more and more human like, something that feels too “unnatural” to some consumers. However, data suggests that millennials have fewer concerns about technology.

• Chatbots had a tremendous growth in 2017 which is why the year is called “the year of the chatbots”. Since then, the use of chatbots has continued to grow.

• The trend of voice-based chatbots is also predicted to grow in 2019.