UBC, Rogers, and MobiledgeX - 5G and Edge Computing Use Cases

Introduction

This post is all about what we learned with Rogers and the University of British Columbia. But first...

Something close to our hearts; we care deeply about finding actual use cases and being prepared to put them in operation. MobiledgeX spends all its time on first discovering the most interesting REAL use cases that are possible now but also will transform the future. It works with the newest devices and the most bleeding-edge developers to understand how to open up their future, and by inference all of our futures, faster. When new thinking meets existing behavior change occurs. We then share all our learning and collaboration into a new initiative called Seamster, where we invite others, perhaps like you, to do the same. We believe a world that is more educated can make better decisions faster and we believe that collaboration and sharing will accelerate that education and learning - for ourselves as well.

The second part of what we do is not the focus of this blog post but I will briefly mention nevertheless. Finding use cases is a first-order problem. The second-order problem is how to easily bring these use cases to the mass market and make them available in a scalable, safe, trusted way. All these new use cases share some very similar attributes. They all involve a very rich understanding of the real world, in most cases via video stream. They all interpret what they are seeing and then respond back with action based on what they understand, and/or send the insights back to a more centralized response. This “real-world understanding” is what is driving edge computing. It is otherwise known as AI and edge computing will democratize its availability. This managing of high data rates with real-time interpretation and response creates a world where the internet starts “to see” on behalf of humans and helps the world to move into a new level of augmented reality. This world is very different from the world we live in already. It requires a whole new level of distribution of enabling microservices in the network to support the devices, machines, and humans they support. And since the devices, machines, and humans are mobile at worse or highly distributed but static, supporting services must be placed in the optimum local proximity to ensure performance and scale is possible. Our CTO Sunay Tripathi recently wrote an introduction to what this means for business leaders of all kinds in Forbes and I recommend you read here - "5G And Enterprise Edge: Developments Toward A Device-Centric View Of The Cloud"

What Did We Learn?

Back to the Rogers UBC hackathon and use case discussion. Why we are here. Ten teams of students participated. All ten teams presented their ideas and where possible their working code. To be fair, the time was short to make working code so there was incredible respect for the teams that showed working prototypes. The ideas and insights were also of premium quality and I want to give great credit to UBC and their student body - the future looks bright. I will summarize clusters of the ideas below and also some of the key insights that were surfaced.

The ability to analyze posture can change controls, rehabilitation, home health.

A cluster of teams understood that being able to watch human form and have real-time analysis of that form can change many situations. One of the teams even went as far as even interviewing a physical therapist on how valuable such a solution could be! One of the problems that restrict recovery is the patient doing their therapy correctly at home. If the quality of the form of the exercises, that are often being done for the first time, is not good then the benefit of the physio is lost and recovery is at best slower and at worse never complete. This was seen to be tremendously valuable for stroke recovery where the brain has to literally relearn basic motion. Another team applied the same logic to improving form in workouts, which will reduce unnecessary injury. Alexander Brock of Rogers, one of the judges, made the observation that this application would be invaluable in work settings where people are lifting heavy objects and are working in places where unusual body positions are required. Health and safety is a large problem for all existing businesses. Another observation is that this form of recognition can be used to monitor people at home to make sure the elderly, for example, are safe and well. Enabling people to continue living in their homes while being safe is a game-changer in terms of later quality of life. The final application that links to this is to enable gesture controls for new ways to control the house systems, enabling people that have limited motion to still control their surroundings. The ultimate control could be the wave to the camera saying “please help me now”.  

Industries impacted: physical therapy, health, and fitness, home care, industrial health, and safety.   

Real-world gaming

There are always gaming ideas. In my experience, the best gaming ideas always take a simple existing gaming experience and move them into a spatial context. One such game was paintball that allowed mass world gameplay, where both active participants and non-participants are part, being sprayed and marked with virtual paint, for different rewards and challenges. The next game was taking Guitar Heroe principles and making an augmented reality version, where rather than playing notes on a plastic guitar, multiple players touch notes suspended above them. Both games highlighted the opportunity to make gaming a social experience that encouraged movement and health. With 5G and edge computing, it was possible to have the same game experience shared over many different makes of devices and also many different generations of devices, removing silos and barriers in the players and also creating a larger addressable market for the game maker. The last insight made was such games were a good enabler for team building and organizational training

Industries impacted: gaming, corporate development

Conference calls you can hear

One brave team took the initiative to not focus visually but to focus aurally. Special kudos to this team since the code samples we distributed tended to be visually centric. Audio augmentation is potentially the most life-changing area of discovery since speech and communication is the foundation of all human interaction. Examples such as real-time aural descriptions of the world for blind people are a game-changer for the 4.1 million visually impaired people in the USA alone. Having universal real-time communication/translation across all languages is no longer a dream but something that will happen in the next 10 years, impacting all forms of society.

The team’s idea is to use machine learning to recognize the voice patterns in any incoming stream, delete the background noise and normalize the volume levels across all conferencing participants. For all people who have been on calls and could not hear other people, either due to a low voice or background noise, this is invaluable. There is a real cost in business to not being able to understand others, from customer support to distributed team interaction. As a side comment.

Industries impacted: Call Center, conference calling, public safety (911).

Small business automated retail security

Large stores can afford security guards, expensive video surveillance systems with live monitoring and large numbers of staff, to attempt to reduce loss due to shoplifting. Small retailers cannot. And all goods stolen are a direct cost to the retailer, that further challenges their ability to compete with online alternatives, where shoplifting is not possible. A simple camera system that can monitor the retail space and automatically detect suspicious behavior is a very cost-effective solution now. A network of small retail owners could also be created where suspicious behavior is shared and the system continuously improves its ability to detect suspicious behavior and people. The small retail owner is enabled to have the same capabilities that previously were only possible for the largest.

Industries impacted: Retail

One Last Observation

One final observation of a common trait across all of these new use cases. Due to the base DNA of using highly rich information of the surrounding area, there is a need for trust and guarantee of privacy. In the same way we all trust making calls on our phones and not having the content of our calls being used and sold without our consent, we need to trust these solutions the same way.  

Thank you to all who collaborated to make the event possible - Rogers, University of British Columbia, Ericsson, Inseego, IEEE Vancouver and of course all from MobiledgeX. From our perspective, we appreciate every opportunity to learn and also contribute. If you would like to execute similar please reach contact us here.