Digital human-assistant framework for distributed applications in 5G cloud networks

5G (ENCQOR) Technology Development Challenge for SMEs

Digital human-assistant framework for distributed applications in 5G cloud networks

Challenge Launch Date:

15-December-2020

Challenge Deadline: 

31-January-2021

Challenge Statement:

5G is a platform to cope with new challenging use cases to exceed the expectations of current human-assistant frameworks. Using the capabilities of 5G such as low latency and very high throughput and smart location of resources will empower digital smart assistants and will bring a new level of perceived Quality of Experience (QoE).

To enable this type of services a new modern and flexible software architecture designed for 5G is required.

Project Partner:

Ericsson Canada Inc.

Timeline:

Up to 12 months starting in 21Q1

Available Funding:

Up to 500 000$ CAD

Applicant Type:

SME with less 500 employees registered in Quebec

Location:

At SME location in Quebec with 5G testing in one of ENCQOR 5G Hubs in MTL or QC

Project Details:

The project goal for the Digital human-assistant framework (DHAF) is to implement a framework designed for 5G which enables different DHA use cases. It shall be easy to adapt for different requirements and shall consider requirements like fast response time, scaling, number of simultaneous sessions, gesture and NLP mapping.

This requires that the software architecture has the capability to adapt easily to different scenarios providing language / sound, video and text-based UI input services that can be consumed by the cloud native applications running in a distributed cloud context possibly in a multi data center setup.

The framework needs to interact with the use case implementing applications and shall be as developer friendly as possible via a generic REST API and high-speed media communication channels. The framework shall be able to generate response options to the user based on User interaction input and history (context / long- and short-term context). The context shall be able to be annotated or enriched by additional services (e.g. an optional location service) for this user interaction. The enriched context shall be considered by the framework when producing responses.

The responses are created with the help of a generic context database. The selection and handling / description of the database is one central part is part of project. It shall be used to map user input intelligent to proposed actions and responses. The database shall be populated/annotated/enriched by applications and ML services which are excluded from this project. A demo application shall be included demonstrating the capabilities of the framework.

To enable mission critical use cases the framework shall be able to execute runtime critical modules or parts of the applications on deep edge nodes possibly residing in a 5G operator network very close to the end user. GPU usage shall be considered for heavy compute operations and ML if available. The framework shall be executed in Kubernetes clusters and follow the microservices architecture design principles. The Software development process shall incorporate privacy and security requirements from start.

Project Goals/Outcomes:

• Deliver a functioning framework in GIT as source code and test code, with a reference setup and HELM deployment charts and associated documentation.
• Readiness to participate in further phases of the product roll out as an extended scope.

Applicant Capabilities:

Experience in

• developing distributed cloud native frameworks and applications, AI/ML, NLP, Video recognition
• Kubernetes, GPU
• “Human to Machine translation” and “Machine to Human translation”
• 5G Networks and 5G use cases

Additional Information:

• Use cases: Smart workspace assistant, Car assistant, Bank assistant, Personal assistant, Shopping site assistant, Digital healthcare assistant, etc.

How to participate:

• Each SME interested to participate to this challenge must have an Encqor iPaaS access (free) in order to understand the 5G technological stacks they will work with for the challenge. SMEs with no iPaaS access need to apply here and select ‘Ericsson Call for projects ‘Digital human-assistant framework’ in the first drop down menu
• Selected SMEs will be invited for a live demo/presentation with the project owners