By Marga Verdú
The Internet of Things (IoT) not only connects computers and mobile devices with all kinds of objects but also has the capacity to connect ‘smart’ cities, buildings, industries, vehicles and homes, as well as infrastructures such as electricity and gas networks and water supply systems, amongst other everyday utilities.
In all these environments, IoT is driving the development of advanced services for processing information in real time, which requires huge volumes of storage space, greater computing power and more intelligent, autonomous sensors.
What features will define this new generation of sensors? What is their operation and efficiency based on? What do they offer for improving the security of processes?
The IoT Solutions World Congress, IoTSWC 2017, identifies some of the key factors defining the adoption of these new sensors, which are being combined with microprocessors and in the short term are set to generate an enormous qualitative leap in the IoT ecosystem.
Efficiency and safety
Scientists Jesús Pacheco and Salim Hariri, members of the National Science Foundation Center for Cloud and Autonomic Computing at the University of Arizona (USA), explain the increased efficiency of sensors through IoT integration as fog and cloud methodologies.
‘In all applications developed for IoT, sensors are the indispensable element to move the physical world to the digital world, a process that benefits from the integration of fog computing.’
According to Pacheco and Hariri, an IoT circuit based on fog and cloud computing not only brings the computational power that these environments demand but is also capable of constructing a new ecosystem of pervasive, cost-effective and fully-accessible services.
The two scientists warn of the security risks resulting from the proliferation of networks of sensors, as they are elements that are vulnerable to certain attacks that are increasingly taking place in Internet of Things infrastructures.
‘However, since fog computing directs services from the edge of the network, its integration into the IoT infrastructure, in addition to optimising the latency and quality of services, has mechanisms that allow the integration of security measures,’ note Pacheco and Hariri.
Meanwhile, American IoT security expert Roopak Venkatakrishnan calls for the use of redundant mechanisms in sensors to detect anomalous responses in terms of security.
Venkatakrishnan says that systems based on redundancy tests are more robust and efficient against sophisticated cyberattacks, compared to the classic reactive defence systems that only work when the attack method is known.
According to this expert, ‘In the world of the Internet of Things, security and anomalous behaviour of sensors and other IoT components has to be determined by more complex commands until there are security components that protect the system based on the sensors themselves that communicate in the real world and the digital world.’
Venkatakrishnan argues that in the IoT world practically any device can have an IP address, an API or a web interface and a certain level of ‘smartness’, where devices are probably software-based, programmable and autonomous.
‘Ideally, one would build security into these devices to make them inherently resistant. Unfortunately, in practice the reliability and security of software-based systems, sensors and IoT elements can be improved only to a certain degree prior to deployment.
Reasons are many, and may range from power consumption restrictions, to complexity, to poor designs or lack of planning.’.
Anatomy of the smart sensor
Majeed Ahmad, the author of ‘IoT and the Evolution of Smart Sensors,’ explains that a window is opening for the optimisation of processing efficiency and data collection in IoT infrastructures with the arrival of MCU, which would be the symbiosis between a sensor and a microprocessor, performing autonomous calibration and diagnosis functions.
This fusion between the sensor and signal processing functions is redefining the sensor landscape, which has traditionally been relegated to discrete elements such as thermocouples and accelerometers.
From the outset, an intelligent sensor allows a signal transfer much closer to the sensor, or even within it, to be sent, which protects signal integrity and ensures the isolation of the IoT system operating in harsh industrial environments.
In addition, a smart sensor can make use of local computing power to process and interpret data in isolation, making decisions based on measurable physical parameters, and establishing the relevant communication.
These types of sensors make decisions about what data to process locally and what other data need to be sent to the cloud through the IoT infrastructure. Smart sensors create boundary conditions without the intervention of a human operator, and their applications can analyse data locally, take action or issue alarms via the cloud to the user in the form of emails, text messages or notifications via the app.
Smart sensors add a higher layer of sophistication to the efficiency of IoT communication processes, and according to Ahmad they have the ability to transform many industrial sectors and everyday environments.
Huge growth potential
Meanwhile, Gartner Research vice-president and partner, David Cearley, highlights artificial intelligence (AI) and machine learning (ML) – which include technologies such as deep learning, neural networks and natural-language processing – as the elements that will support the most advanced platforms.
These will include sensors with the ability to understand, learn, predict and adapt and potentially operate autonomously. ‘Systems can learn and change future behaviour, leading to the creation of more intelligent devices and programmes.
The combination of extensive parallel processing power, advanced algorithms and massive data sets to feed the algorithms is unleashing this new era.’
According to Cearley, the lines between the digital and physical world continue to blur, creating new opportunities for digital businesses.
The man known as the ‘father of sensors’, Janus Bryzek, argues that there are multiple factors accelerating the mass adoption of IoT ecosystem: ‘First, there is the new version of the Internet Protocol, IPv6, enabling an almost unlimited number of devices connected to networks.’
The expert adds that another key factor is that major network providers such as Cisco, IBM, GE and Amazon have decided to support IoT ‘with network modification, adding Fog layer and planning to add Swarm layer, facilitating dramatic simplification and cost reduction for network connectivity.’
As a consequence, Bryzek forecasts that ‘with GE estimating that the Industrial Internet has the potential to add $10 to $15 trillion to global GDP over the next 20 years, and Cisco increasing to $19 trillion its forecast for the economic value created by the Internet of Everything (IoE) in 2020), the IoT ecosystem will experience the largest growth in the history of humankind.’