Olly the Robot: The Next Big Home Assistant?
A project by Emotech and Heriot-Watt University has developed an innovative new robot, Olly, that marks a significant step in how conversation interfaces can be applied across various industry uses.
AI technology specialists, Emotech, had grand aspirations to develop the world’s first robot with personality; one which can identify specific users and finely cater to their individual needs and preferences. That vision has become a reality following a collaborative project between some of Scotland’s finest innovation hubs.
Emotech worked with Heriot-Watt University and was supported by The Data Lab to develop the robot’s natural language processing abilities. The Data Lab’s support and funding for the project has since allowed the AI specialists to research and develop test models; resulting in Olly.
Olly can perform most of the standard tasks that one would associate with home assistants. It can play music, dim lights, write emails and set alarms. However, Olly is a unique creation and is, by all means, on another level compared to its counterparts. Its ability to understand and generate language is groundbreaking and allows it to remember specific user requests and anticipate needs with staggering efficiency.
Idea to Reality
Dr Pawel Swietojanski, a research scientist at Emotech, worked on the project since its launch in December 2016. Dr Swietojanski said that after meeting Professor Verena Rieser and Dr Xingkun Liu from Heriot-Watt, they quickly realised there was a groundbreaking opportunity to collaborate on the Olly project.
He said: “When I met Dr Rieser and Dr Liu from Heriot-Watt’s School of Mathematical and Computer Sciences at a natural language generation conference we realised that there was an opportunity to work on this project together.
“The Data Lab was instrumental in making that happen by helping fund Emotech’s collaboration with Heriot-Watt and therefore making Olly a reality.”
Indeed, The Data Lab’s funding proved crucial to this project, funding 50% of the project costs and enabling the tie-up with Heriot-Watt University’s Interaction and Natural Language Processing Lab – led by Dr Liu and Dr Rieser. This collaboration allowed research to be undertaken, leading to the development of Olly’s unique framework and test model.
How Does Olly Work?
Olly is both a showcase of Emotech’s AI technology and represents a critical first step in demonstrating how conversation interfaces can be applied across industry uses.
Dr Liu of Heriot-Watt University said: “AI is an increasingly popular topic but often people are unclear on how it can be used. This project demonstrates how the technology can be applied across different industries.
“For instance, we are currently researching how the research-based, data-driven NLG can be adapted to industrial products such as Olly. It’s great that The Data Lab is facilitating this type of research in Scotland.”
It uses camera sensors that can establish how many users are in a room at any given time. Once Olly is aware of who is present, it then adapts its responses to the user or group.
If a user were to speak, the robot then used automatic speech recognition (ASR) to convert the users’ speech into text utterances. Natural Language Understanding (NLU) follows ASR and is how Olly learns each user’s preferences.
Heriot-Watt University developed Olly’s NLU through open source software, Rasa_NLU, while Emotech collected and compiled real user data to train the NLU component. For example, if a user says “please set my alarm for 6am”, Olly’s NLU component will understand the user’s intention to set an alarm.
Data-driven NLU has a number of notable advantages over rule-based NLU, because the robot in question is capable of learning from real user data and can, therefore, better respond to real situations. This data-driven approach is also easily extended and adapted to new domains.
Upon completion of NLU, Olly has a Dialogue Manager (DM) which enables it to determine what to do, or say, and learns from previous conversations to remember a specific user’s intentions or preferences. Natural Language Generation (NLG) is also a key component in Olly’s processes, however, this is currently template-based according to Emotech.
Finally, Olly uses Text-To-Speech (TTS) to convert the natural language text to speech which the user will hear.
Olly the Robot could become a household name in years to come, providing users across the globe with intuitive, real-time assistance. The product is available from September and will be available on Indiegogo for around £415.