Ending the data deadlock: why Industry 4.0 will require us to change the way we think of managing big data
By Adam Byrne
By Adam Byrne
September 13, 2018 – Recent reports project that Canadian industry is perfectly positioned to take advantage of a new industrial revolution and the process is already taking place.
A survey conducted by PwC revealed that over half of Canada’s industrial companies expect to achieve advanced levels of digitization within two years and that 31 per cent have already passed this milestone.
It is envisaged that industrial digitalization will cause a shift in production, similar to that of the invention of mechanization, creating a digital industrial supply chain where products are developed, manufactured and monitored in real-time from design studio to display case through a single, integrated process. Connected supply chains with smart sensors will allow the real-time monitoring and measurement of vast automated assembly lines.
The possibilities are huge. The digital blurring of the boundary between the physical and virtual world in manufacturing will also allow computer visualizations to be seamlessly converted into customized physical products in any location through 3D printers and other machines. Industrial robots capable of manufacturing everything from aircraft components to nuclear parts will create cost-effective, fast, efficient and low-maintenance industrial supply chains.
Yet the road ahead is replete with major obstacles and challenges. According to Deloitte, 87 per cent of Canada’s businesses report they are not ready for the disruption that advanced technologies will create, and 32 per cent consider themselves completely unprepared.
Central to the vision of Industry 4.0 is the creation of digitally interwoven automated supply chains that can continuously adapt to live-market trends or trading conditions through the ability to track, trace, monitor and respond in real time. This will require a revolution in industrial communications where big data can be digitally omnipresent across a supply chain the second it is generated, just as remote access technology currently allows the data on one screen to be replicated in real time on another. We will need new open standards to enable seamless interchange of data across the industrial ecosystem.
This will be essential to enable control signals, sensor measurements and other data to be continuously exchanged in real time so that all critical industrial machinery, from robots to 3D printers, can be instantly accessed remotely and controlled by humans.
With robots making safety-critical components such as aircraft engines, staff will need to be able to remotely monitor, optimize and maintain them to instantly detect and fix software faults or prevent malicious interference. Cybersecurity teams will also need to be able to remotely monitor and intervene in manufacturing machinery to avert industrial sabotage by cyber-attack. For example, researchers have previously shown how to hack into 3D-printer files to make a drone crash by altering the design specifications for the propellers.
There are also plans to radically decentralize manufacturing through an “Internet of Thinking” based on a “DIY” sensor network that can autonomously analyze information, rather than send it for remote analysis. This will mean that if an industrial robot notices its own equipment malfunctioning, it will be able to recognize what it needs to do and self-correct any faults. However, such systems will need external human supervision in case of faults that they fail to detect or if they involve errors that require human intervention.
Continuous live data will also be critical to enable armies of robot “workers” to operate in seamless synchronicity, continually choreographing their actions in response to live location or measurement data from other robots.
The need for data to be digitally present in multiple places at once the moment it is generated requires us to apply the thinking behind the traditional “IT help desk” to the whole of the industrial IoT ecosystem. We need a secure standard for sharing everything from audio and video to images and text among millions of industrial devices. This will require a genuinely low-latency, live connection that can make industrial control rooms digitally omnipresent across an entire supply line by allowing the two-way exchange of control signals and information.
Yet there are currently no open standards for industrial remote communication to allow industrial machines from any vendor to share live data and enable remote human intervention across all manufacturing equipment and components. Companies have proprietary protocols for remote access to machinery, but these work only with their own machines. This proprietary model for machine-to-machine has arguably damaged other industries.
Millions of vehicles do not connect with different phone models, thus creating a fragmented “connected car” IoT due to car dashboards being divided between rival smartphone firms that “lock out” their competitors’ devices.
If this system is transposed into manufacturing, it will create a fragmented industrial IoT landscape where some machines cannot interface with machines by other manufacturers and control rooms cannot exercise the same degree of control over all machinery. This undermines the vision of digitally integrated end-to-end supply chains – and it could also make automation impossible, as it would necessitate real-time remote supervision of all machinery. Crucially, it could jeopardize cybersecurity by making some machines inaccessible to cybersecurity experts and by creating many different attack vectors that are difficult to protect.
The only way forward for manufacturers is to create a safe, secure open platform for end-to-end data exchange across the entire industrial supply chain. This would enable low latency, multi-directional communication among all machines. Significantly, it would mean that supply-chain connectivity is backwards compatible and fully “future-proofed,” so it can seamlessly incorporate any new industrial robots or machines that emerge in the future. It would also make technicians, cybersecurity personnel, engineers and factory staff digitally ubiquitous across a factory floor of diverse and varied equipment, and enable them to instantly remote into any machine to fix faults. If we are to fully realize the vision of an interconnected industrial supply chain across Canada and the rest of the world, we must challenge rival industrial manufacturers to convene around a more inclusive open-standard approach that will strengthen all the stakeholders in the supply chain.
Adam Byrne is CEO of RealVNC. With a background in mathematics and computer science and a career spent growing IT companies from the ground up, he is building a worldwide ecosystem of technology partners in a variety of vertical markets.
This article originally appeared in the September 2018 issue of Manufacturing AUTOMATION.