As the number of connected and smart devices increase, we are now entering a world where everything will be measurable and can be analysed. The next step in digital transformation related to Internet of Things (IoT) technology will be 'digital twins' as we enter a world of data intensive science.
A digital twin is a virtual software replication of a physical asset. It has all the components of the asset and can help understand the dynamics of how the asset operates. Originally, the concept mainly focused on machines, but has now been expanded to include smart buildings, smart cities, telecommunications networks, human body, farm crops and business processes.
Speaking on how digital twins will impact the future of IoT analytics, Stephen Brobst, Chief Technology Officer, Teradata Corporation said, "We can take our theories and our formulas and now measure the real world, and instead of theorising about what's going to happen, we can measure and translate it into the analytics and predict and prescribe what are the next steps and action to take.
The digital twin products that are deployed have the ability to self report on what's going on with them. "With digital twins we are not just delivering objects that do things but delivering the ability to do analytics and create better business outcomes. Digital twin is the next step in advanced software capabilities and advanced analytics in order to realise those outcomes. There is a lot of development still going on in this space. It is a very active area of research," Brobst highlighted at the Teradata Universe Conference in London.
Teradata is working closely with companies like Siemens, Cisco and GE to create these capabilities and partnerships to leverage the Teradata Analytics platform into this world of digital twins.
Digital twins can also work in bringing about disruption by not just transforming physical objects into digital ones, but also vice versa. A digital twin takes the digital transformation concepts one step further by linking the world of atoms and bits through use of digital copies of a physical object that co-exist in time. "We are working with Cisco on an animal rescue digital twins project as we design flippers for a tortoise that had to be amputated, to allow him to swim again. Digital twins can help us simulate, predict and ultimately prescribe the actions that we should take to create the best possible outcome," Brobst explained.
The concept of digital twins has been around since 2002, but it is only recently that it is getting enabled thanks to IoT technology. "With the third wave of Big Data focused on sensors and Internet of Evertything (IoE) the concept of digital twins becomes truly real as we now actually have the data," he added.
Within the next five years, it is predicted that 85 per cent of IoT platforms will be using digital twins technology. "The outcome of deploying this technology will be 25 per cent improvements in productivity and efficiency. Digital twins can take those software proxies of an object or a system or a process and predict outcomes and use the analytics to prescribe actions," Brobst opined.
According to a recent Gartner survey, 48 per cent of organizations surveyed that are implementing IoT are already using digital twins or are planning to do so in the next year. "A digital twin is powered by the analytics and data is the fuel that empowers the analytic algorithms. We can use this capability in all stages of the lifetime of an object from design to building, to deploying and to operating," he said.
In Singapore, there are sensors collecting data on taxis and the network transportation data. "Local bodies can start small and then can grow that data incrementally, with more data and more advanced analytic capability. Pick an area where an opportunity is greatest for your city - be it healthcare, waste management or transportation. Once we track them we will start understanding the interaction between all these systems and how to optimize in your digital environment," he added.
But the reality is that less than one per cent of sensor data that is being collected is actually used for creating value. "This is a big opportunity. We can use that data in a much more sophisticated way than what we are doing today. We need to move away from the idea of very unsophisticated rules engines which are encoded by human subject matter experts to a world in which we are continuously learning using machine learning technique," Brobst explained.
McKinsey Global Institute (MGI) estimates that approximately 70 per cent of value creation from IoE will be in the B2B domain. "This is a very big leapfrogging opportunity for our traditional manufacturing customers. The value propositions for your businesses is going to change, pricing models will change. We are not going to see more hardware, but better software. It is the algorithms and the software that will drive the value in the next generation of manufacturing, process design. Businesses should rethink what that means in your investment in analytics," he concluded.
The journalist was hosted by Teradata at the Teradata Universe EMEA 2018.