FountainBlue’s August 9 VIP roundtable, on the topic of ‘Innovating on the Edge’. Please join me in our gracious host at Intel and to each of you for your input and advice. Below are notes from the conversation.
There are many factors which lead to the emergence and growth of Edge computing, including:
- the volumes of raw data generated by the exploding number of devices, processes, and programs;
- the complexity of data available makes it more difficult to process, filter, understand;
- the users are demanding powerful, personalized, complex solutions which are secure and private;
- the imminent arrival of 5G solutions will push executables down faster to the ‘edge’, the device itself;
- the tremendous need for energy and power (and the associated expense) if everything is processed on the cloud itself;
- the urgent need for quick responses, especially when safety and lives are on the line; and
- the immediate and ongoing need to protect the privacy of users, the security of systems and devices and networks.
But it’s no easy task to innovate on the Edge.
- Each solution must be able to efficiently filter out data, focusing on the ‘real’ data, the ‘relevant’ data for the problem at hand.
- There’s a challenge to make strategic decisions around technology and business, while also not getting stuck with the decision made, in case things don’t go as planned.
- The current investment environment is pro-software and less bullish on hardware in general.
- Each solution must navigate the technical, business and regulatory objectives and constraints, while also solving the problem.
- The speed of change is mind-boggling, and innovating in that environment is difficult at best. But things also keep evolving and changing, which makes things even more difficult.
- Memory and storage bottle necks may arise with the rise in volume and complexity of data and processing.
- It’s a sobering thought, but devices and solutions on the edge which might be turned into weapons (including cars) have additional security and operational requirements.
- Companies must also protect itself from financial, legal and brand exposure should a solution on the Edge cause unintended damage to users.
Below are some thoughts on how to keep that innovative edge.
- Be strategic.
- Know what your customers need in the short term and for the long term and plan accordingly.
- Work with an ecosystem of partners to deliver tailored solutions efficiently.
- Adopt a set of Open Source tools which would help rapidly develop, deploy and manage solutions on the edge.
- Develop hardware-agnostic solutions which are more versatile and adaptable.
- Adopt self-maintenance systems to ensure validity of solution and ongoing maintenance. With that said, do not delegate on management to automation. Know when the scenarios when you need proactive leadership and management and respond accordingly.
- AI will take you far – understanding the relevant data. ML can take you farther – it could help you understand the trend and make predictions beyond the historical data. Both are necessary and essential. Progressively more of the AI will take place closer to the edge.
- The speed and accuracy for data processing is essential for Edge Computing, as it is for just about everything else involving data. The ability to process unstructured data and video and the ability to focus on the deltas rather than the raw data will help solutions better manage and filter data.
- Collaboration is key as there are so many players involved.
- Carriers need to invest in 5G.
- Cities need to adopt the infrastructure for 5G.
- Each solution is a combination of hardware, software, processes, etc., Partnering with others in non-core offerings is essential.
- Privacy and security must be maintained. Having ecosystem partners focusing on these areas will help companies focus on delivering on their core value.
In conclusion, our leaders agreed that innovators in the Edge Computing space must create an ecosystem of players and connect with players across the ecosystem at many levels.
You must be logged in to post a comment.