In the third article of the series based on the panel AI On The edge vs the cloud that took place during the Industrial AI Summit and with Alex West – Principal Analyst at Omdia, Matteo Dariol – Lead Innovation Strategist at Bosch Rexroth, Anders Rahm-Nilzon – Director of Cloud centre of excellence at Volvo Group, and our CEO and Data Science Director Eric Topham, we discuss what are the elements that are impacting the future development and deployment of industrial AI.
What is impacting the future development and deployment of industrial AI?
Data handling is massively impacting the industry at the moment. Most businesses want to keep their data within their premises and this has the advantage of them being able to look over the data that they have, and also being able to access data directly to create those networks and solutions that they want to deploy.
In reality, with the current state of technology, with the right tools and the right connectivity stack in place, it doesn’t make too much difference to access a local database versus a cloud database and manufacturers and industrialists are realising this.
Moreover, scalability and flexibility should be the decisive factors in their decision-making process and the cloud will hugely help.
Nowadays, you can prototype things using small compute instances, and seamlessly scale up onto larger instances, and then scale back again, once a particular training run is done. This will help manufacturers to observe if the project will deliver what they wanted and help them allocate the right amount of resources to it.
The phase on which manufacturers find themselves is also having an impact. Depending on if they are in their development cycle or a deployment cycle will set the scene for what comes next for them
Ultimately, portability is now becoming an additional key element that is impacting the future development and deployment of industrial AI. How easy it is for organisations to select something and use it agnostically, either on the edge of the cloud, will impact the future of the industry
A lot of industries are trialling projects. In some cases, the Proofs Of Concept (POC) are successful, but even then, there is still lots of work to do to move from successful POC to successful deployment. The industry is moving towards a scenario where there isn’t much of an issue in scaling AI as a technology, but scaling the organization is much harder. And organisations are now starting to realise this.