Periodic Reporting for period 1 - Totem Spoon (Interactive Digital Signage with emotional intelligence for smart cities)
Période du rapport: 2019-06-01 au 2019-11-30
From connected cities to connected buildings, malls to museums, private or shared vehicles to public transports, all these platforms share the same need: to welcome people, exchange, assist, provide the relevant information and connect them with the right service. More and more digital services are becoming available in these collective places, through smartphones and other digital devices. Accessing information has never been easier. However, public places are struggling to push relevant information to their users, due to the multiplication of applications to install, and the lack of proper interaction provided by the current digital signage generation... Voice only interfaces, such as connected speakers, despite their multiple advantages (efficient at realizing complex tasks, hands-off interaction makes it inherently health-safe to use, …), are not adapted to public spaces either. Without an embodiment, these interfaces are neither proactive and self-explanatory, nor intuitive enough for people to use comfortably.
With the prominence of voice interactions - in a recent market research report by Microsoft, 72% of survey respondents reported using voice commands, and 66% of those are using it at least once a week - the information world is changing, driven by the Millenials. Public spaces need to adapt to this new paradigm without enlarging the digital divide. SPooN makes it possible for the locality to push information to its users instead of relying on unverified news channels. By bringing these services to the front, SPooN is an enabler for the AI ecosystem in public places. Driving the adoption of connected services will generate more users and drive the economy in this local ecosystem. Furthermore, the GAFAs are not occupying these public places. It is a unique opportunity to deploy technology with strong European values, based on our Ethics by Design approach.
SPooN has developed “SPooNy”, the next generation of interactive digital signage by transforming existing digital signage equipment installed in public spaces (city centers, airports, stations, …) into interactive and marvelous creatures able to interact intuitively with people. The overall objectives of this SME Phase 1 are to validate our technological approach and the product-market fit. During this project, our main focus is on numerous deployments that will drive our understanding of the market while validating our approach. It will also be complemented by market research.
With the prominence of voice interactions - in a recent market research report by Microsoft, 72% of survey respondents reported using voice commands, and 66% of those are using it at least once a week - the information world is changing, driven by the Millenials. Public spaces need to adapt to this new paradigm without enlarging the digital divide. SPooN makes it possible for the locality to push information to its users instead of relying on unverified news channels. By bringing these services to the front, SPooN is an enabler for the AI ecosystem in public places. Driving the adoption of connected services will generate more users and drive the economy in this local ecosystem. Furthermore, the GAFAs are not occupying these public places. It is a unique opportunity to deploy technology with strong European values, based on our Ethics by Design approach.
SPooN has developed “SPooNy”, the next generation of interactive digital signage by transforming existing digital signage equipment installed in public spaces (city centers, airports, stations, …) into interactive and marvelous creatures able to interact intuitively with people. The overall objectives of this SME Phase 1 are to validate our technological approach and the product-market fit. During this project, our main focus is on numerous deployments that will drive our understanding of the market while validating our approach. It will also be complemented by market research.
SPooNy is deployed at EPOKA in a company building, in order to offer a complete welcoming experience, next to the welcoming desk. SPooNy is allowing visitors to send a notification from visitors to the visited person in the building, choosing from the company listed employees. Other services are under development (meeting room booking, company news, services when exiting the building …). The technical solution developed during this project can be reused for different companies.
SPooNy has been live for 1 month in November 2019 at a Mall Center. The aim for Toyota services branch was to deliver a welcoming & drive to store operation to its customer Uniqlo, in a big mall / commercial center in Tokyo. SPooN was delivering an entertaining experience, specially dedicated to kids. The aim was to drive traffic towards the Uniqlo shop.
SPooNy is deployed in a Toyota car dealer shop in Tokyo for a minimum duration of 6 months. SPooNy welcomes new visitors, and engage with them. In step 1 of the experimentation (Dec 19 - Feb 20), services are aimed at welcoming and addressing clients to the shop team. In step 2 (March 20 - 20), SPooNy will deliver new services, provided by Toyota, which will be plugged within our ecosystem.
SPooNy has been live for 1 month in November 2019 at a Mall Center. The aim for Toyota services branch was to deliver a welcoming & drive to store operation to its customer Uniqlo, in a big mall / commercial center in Tokyo. SPooN was delivering an entertaining experience, specially dedicated to kids. The aim was to drive traffic towards the Uniqlo shop.
SPooNy is deployed in a Toyota car dealer shop in Tokyo for a minimum duration of 6 months. SPooNy welcomes new visitors, and engage with them. In step 1 of the experimentation (Dec 19 - Feb 20), services are aimed at welcoming and addressing clients to the shop team. In step 2 (March 20 - 20), SPooNy will deliver new services, provided by Toyota, which will be plugged within our ecosystem.
For this project, our aim was to test SPooNy’s interactions with passers-by, consumers/prospects in order to validate the pertinence of our solution for different business cases. We wanted to focus on deploying in real conditions, in diverse locations, and with different scenarios and use cases, in order to test a wide panel of uses for SPooN’s technologies. That’s why we targeted 5 different deployments (using 4 SPooNies).
The outcomes we focused on during those deployments were:
- to gather quantified data, consumer feedback, understand the interactions, needs and requests from consumer on SPooNy in order to influence the roadmap
- understand feedback from companies & corporations working with us on those tests
- gather all needs, assess & identify business markets, validate and amend our current offers, gather the information we need in order to reach mass market with proven results & feedbacks
The work for SME was based on our existing work at SPooN. Our products were already deployed in different situations (retail with LVMH, Hôpital Broca, le Cent Quatre, ...) and we improved our products during the project focusing on two main axes: developing pertinent scenarios given the use case we wanted to study, and improving our analytics platform to get more meaningful data out of the experiments.
Technical focus 1/ Various scenario development:
As described in the following document, we have deployed SPooN in a variety of locations during the span of this project, to solve real-world use cases. Those deployment needed some custom scenarios development: using the tools that we have built internally, the behavior of the Spoonies was tailored to fit the given situation and solve the given use case. The new developments are highlighted in section 1c.
Technical focus 2 / SPooN Analytics:
A great part of the technical work during this SME project was focused on SPooN Analytics. SPooN Analytics is the live analysis solution for our deployments. The goal is to be able to remotely track the quality of the interaction of our solutions in order to have an objective measurement of our performance. This is crucial both for ourselves, in order to keep improving our services, and for our clients or future clients, in order to justify the usefulness of a deployment.
The solution we selected is to deploy an Elasticsearch-Logstash-Kibana stack (usually referred to as the ELK stack) in the cloud on a Kubernetes cluster. This widely used open source stack is a standard for extracting actionable insights from a vast amount of data.
Using this solution, we can easily generate a report for our clients on the usage of the robot. This report contains easily readable graphs picturing some functional data on the usages, such as:
- How many people did interact with Spoony today?
- How many scenarios where launched, how many times did they last and how did they end?
- Which scenario is more appealing to a certain gender/age category?
- Is there a prefered interaction mode? Does it evolve with gender or age?
- What do people ask the most? What does Spoony doesn’t understand?
- Which scenario choices are prefered per category?
We also use the solution internally to monitor our deployed robots, and are using even more functionalities than the data provided to the customers. For example, we monitor so that we can react as soon as it happens to fix the problem. We also use it to generate internal indicators - such as the number of minutes that our robots have been interacted with when they are running. This is a good indicator of the attraction of our solution and we focus our efforts on maximising it.
The outcomes we focused on during those deployments were:
- to gather quantified data, consumer feedback, understand the interactions, needs and requests from consumer on SPooNy in order to influence the roadmap
- understand feedback from companies & corporations working with us on those tests
- gather all needs, assess & identify business markets, validate and amend our current offers, gather the information we need in order to reach mass market with proven results & feedbacks
The work for SME was based on our existing work at SPooN. Our products were already deployed in different situations (retail with LVMH, Hôpital Broca, le Cent Quatre, ...) and we improved our products during the project focusing on two main axes: developing pertinent scenarios given the use case we wanted to study, and improving our analytics platform to get more meaningful data out of the experiments.
Technical focus 1/ Various scenario development:
As described in the following document, we have deployed SPooN in a variety of locations during the span of this project, to solve real-world use cases. Those deployment needed some custom scenarios development: using the tools that we have built internally, the behavior of the Spoonies was tailored to fit the given situation and solve the given use case. The new developments are highlighted in section 1c.
Technical focus 2 / SPooN Analytics:
A great part of the technical work during this SME project was focused on SPooN Analytics. SPooN Analytics is the live analysis solution for our deployments. The goal is to be able to remotely track the quality of the interaction of our solutions in order to have an objective measurement of our performance. This is crucial both for ourselves, in order to keep improving our services, and for our clients or future clients, in order to justify the usefulness of a deployment.
The solution we selected is to deploy an Elasticsearch-Logstash-Kibana stack (usually referred to as the ELK stack) in the cloud on a Kubernetes cluster. This widely used open source stack is a standard for extracting actionable insights from a vast amount of data.
Using this solution, we can easily generate a report for our clients on the usage of the robot. This report contains easily readable graphs picturing some functional data on the usages, such as:
- How many people did interact with Spoony today?
- How many scenarios where launched, how many times did they last and how did they end?
- Which scenario is more appealing to a certain gender/age category?
- Is there a prefered interaction mode? Does it evolve with gender or age?
- What do people ask the most? What does Spoony doesn’t understand?
- Which scenario choices are prefered per category?
We also use the solution internally to monitor our deployed robots, and are using even more functionalities than the data provided to the customers. For example, we monitor so that we can react as soon as it happens to fix the problem. We also use it to generate internal indicators - such as the number of minutes that our robots have been interacted with when they are running. This is a good indicator of the attraction of our solution and we focus our efforts on maximising it.