subscribe: Daily Newsletter

 

Disruptive technologies to look out for

0 comments

Kathy Gibson reports from Gartner Symposium in Cape Town – Disruptive technologies are driving the new world of algorithmic business, with digital mesh and smart machines enabled by new advances in IT.

Brian Burke, research vice-president at Gartner, says the 10 technologies that will drive change fit into these three categories.

The digital mesh refers to the explosion in devices including 3D printing and impact on organisations, and includes:

  • Device mesh – all of the device explosion of devices around us. By 2020 there will be 25-billion devices in the mesh, Burke says. “That explosion includes things like smart TVs, connected automobiles, connectivity in planes and sensors everywhere. The question is how are you interacting with them? You will interact with multiple devices that are surrounding you wherever you are – and a changing mix of devices that will be working concert with your applications and other devices,” he says. “The devices around you will be aware and available to you. The idea of mobility is changing to a mix of devices that are around you and how applications interact with them.”
  • The ambient user experience – being able to integrate the devices that surround you – be they yours or others. “It will be a contextual world in terms of how devices integrate with our needs,” Burke says. “This will necessitate changes in application design. The applications will be contextualised for the devices around you.”
  • 3D printing materials – more particularly, the new and lower-cost materials that are available. “The interesting thing this year is the explosion of new materials,” Burke says. “For instance we can now print using calcium phosphate so you can print bones. We can do 3D printing of organs, skin and other things in the medical area, as well as in aerospace; and actually printing in space. Where we are going is towards composite materials, where 3D printers can print simultaneously with various types of materials.”

Smart machines include autonomous vehicles, different approaches to learning machines and autonomous agents and things:

  • The information of everything – all of the 25-billion devices that will be connected will generate enormous quantities of information. “It becomes a huge information management challenge,” Burke says. “What do you keep, what do you throw away, what do you analyse? That data is not just structured but all kind of data, and it all has to be managed. A key issue will be for organisations to understand what is the most important information and to manage that.”
  • Machine learning – computers today follow instructions, a learning machine has an objective and using different approaches it is sorting through a lot of information to uncover patterns. “These machines are insightful and can uncover patterns that humans aren’t able to,” Burke says. “They do this by going through a lot of material and proposing solutions. There is a lot of development and it will impact all of these other trends. We will have computers that are getting smarter and able to do things the surprise us, things like image recognition and natural language processing where we will see tremendous improvements. We are moving to machines that are adaptive, insightful even curious.”
  • Autonomous agents and things – these are smart advisors; virtual customer assistants; and virtual personal assistants. Smart advisors like Watson are learning about a particular cancer and making suggestions for treatments. Virtual customer assistants understand a question, look through the database and learn over time what are the correct answers. “Virtual personal assistants are a very exciting area; they will act like your personal assistant in many ways to manage your mail, correspondence and meetings,” Burke says. “We will get to the post-app world where your VPA will interact with apps on your behalf.” Meanwhile, autonomous things are beginning to act on their own behalf. “Regulation is holding them back at the moment,” says Burke. “There are lots of autonomous vehicles in controlled environments now, but in uncontrolled environments – on the roads for instance – we are moving forward. In part, this technology is advancing really quickly; and, in part, regulation is holding it back.”

The new IT reality refers to the underlying technology and infrastructure that supports these trends, and includes changes in security architectures, application design, and Internet of Things (IoT) platforms:

  • Adaptive security architecture – the requirement for security is shifting. “The explosion in devices increases the threat surface; we also increase the complexity we deal with,” Burke says. “So we have to change the way we think about security from block and prevent to a more proactive approach where we are detecting issues, responding to them, integrating that into applications throughout the development, testing and production processes. We have to analyse both user and entity behaviour.”
  • Advanced systems architecture – this is mostly in the hardware space. “Driven by smart machines, shifts in hardware in terms of raw compute power will leverage GPUs (graphic processing units) in combination with CPUs (central processing units) because a lot of the processing is pattern resolution that can be done in parallel,” Burke says. GPUs have seen massive processing power increases over the last few years, he adds. “Field-programmable gate arrays (FPGAs) are starting to be used in smart machine technologies because they consume low power and can be programmed in the field. All these FPGAs and different types of hardware, such as neuro-morphic chips, are trying to basically mimic neural networks in our brains.”
  • Mesh app and service architecture – involves using containers like Docker to virtualise applications at operating system level; and microservices which can be dynamically integrated into an application. “We need to modernise application design using containers and virtualising the operating system,” Burke advises.
  • Iot architecture as a platform – IoT platforms will be needed to manage things, to transmit data, to provide a security mechanism and data storage. “At the moment we have a very fragmented market in IoT with a real mix of capabilities,” Burke says. What would be ideal is a general-purpose IoT platform but that’s not what we have today. But some big players are entering the market and we expect to see a lot of market churn until a couple of players emerge who will dominate with an IoT platform.” He adds that an IoT platform is crucially important to the future of IoT, largely because of the security issues raised by so much data.