Real-time software is becoming the cornerstone of intelligent robotics, enabling a new era of safe, high-performance automation. As global manufacturing faces rising complexity, only those with the right digital foundation will keep pace.
Europe’s manufacturing sector is no stranger to disruption. It has endured profound structural change, from pandemic-induced shocks to skilled labour shortages and global supply chain volatility. Yet amidst these pressures, one truth has emerged with greater clarity: automation is no longer a luxury. It is a strategic necessity.
However, automation alone is not enough. Real productivity gains are increasingly driven by the convergence of robotics and intelligent software, particularly in the spaces where precision and safety must coexist in real time. According to Winston Leung, Senior Product Marketing Manager at BlackBerry QNX, the shift now underway is not simply about machines replacing people but about reconfiguring the manufacturing environment to operate more efficiently and flexibly.
“There is a clear transition occurring in how robotics are deployed,” Leung explains. “Previously, we designed spaces where robots and humans were separated, fences, laser sensors, and dedicated zones. Now, we are seeing far more unstructured environments. Robots are working in collaborative spaces, sharing physical areas with people. This transformation is at the heart of what some call Industry 5.0.”
The implications go beyond logistics. Collaborative robotics demands a new tier of technical capability. It requires intelligent systems that respond precisely, navigates unpredictability, and guarantee functional safety under variable loads. The enabler of this evolution is not the robot itself but the foundational software embedded deep within the control architecture.
Real-time performance defines manufacturing agility
As manufacturers adopt a wider range of robotic systems, from articulated arms to autonomous mobile robots (AMRs), the requirements placed on the underlying software grow exponentially. Consistency, responsiveness and fault tolerance are no longer competitive differentiators; they are minimum entry requirements.
The real-time operating system, or RTOS, is at the centre of this reliability. BlackBerry QNX, a longstanding provider of RTOS and hypervisor technologies, has embedded its software into a wide variety of industrial control systems. “We work closely with silicon partners like AMD and Intel,” Leung says. “For example, our collaboration with AMD enables support for their Korea robotics platform, while our partnership with Intel helps bring functional safety to robotics and industrial automation.”
The performance of the software stack is essential. Low latency and low jitter, how quickly and predictably the system responds to an input, directly impact production precision. “When we talk about precision and repeatability in automation, what we really mean is ensuring the system can deliver the same outcome every time, at sub-millisecond intervals,” Leung continues. “Every signal must be received, processed, and acted upon within a tightly controlled timeframe. That requires foundational software to guarantee real-time behaviour under all operating conditions.”
This focus on timing and predictability becomes even more critical when robots operate in proximity to human workers. In this context, software must also support safety protocols that respond instantly to environmental triggers, such as a human stepping into a robot’s path or a sensor detecting an unexpected obstacle.
Safety and complexity are escalating in tandem
Automation systems are growing more complex, with AI workloads, vision systems, and multi-robot coordination placing additional demands on processing infrastructure. Manufacturers that previously relied on microcontroller-based architectures are increasingly adopting more powerful MPU systems to handle dynamic workloads and real-time analytics.
This is where architectural resilience matters. The QNX microkernel design ensures that faults in one software component do not cascade into system-wide failures. “Unlike monolithic kernels, where a driver fault might bring down the entire system, our architecture allows failed components to be restarted independently,” Leung explains. “This is vital for applications where downtime is unacceptable and functional safety cannot be compromised.”
The growing adoption of AI within robotic platforms also brings new challenges. As machines begin to process more data, make local decisions, and adapt their behaviour in real time, software must be fast and dependable. “AI is just another workload,” Leung adds. “But it is a workload that depends on the speed and consistency of the underlying software. The faster you can provide processed data, the quicker AI can respond, directly impacting how safely and effectively the robot operates.”
The movement towards collaborative robotics is also reshaping regulatory standards. The ISO 10218 update, expected next year, will introduce new safety requirements for human-robot interaction, including applications such as speed and separation monitoring, haptic feedback, and force limitation. These requirements are only achievable with software that can handle rapid, multi-source data inputs and deliver deterministic responses without fail.
Embedded from the ground up
BlackBerry QNX does not sell robots or complete systems. Instead, it operates several layers beneath the visible surface, embedded within the robot’s control system, motion controller or safety module. Its software is often selected during the early stages of design as robot OEMs define the performance envelope of their next-generation systems.
“If you are talking about an articulated robot arm, we are embedded in the motion controller,” Leung explains. “Our software sits directly on the processor, with the robot’s control software running above it. This applies to mobile robots and surgical systems, although that falls outside manufacturing.”
The implications for manufacturing operations are clear. QNX’s involvement begins long before a manufacturer purchases a robotic solution. For system integrators or end users deploying robots into the production environment, the performance of the underlying software is already baked into the machine’s capabilities.
Leung is direct when asked what initial steps manufacturers should take to deploy such software within operations. “This is not a packaged product where you install it and go,” he explains. If you are Amazon, Boston Dynamics, or similar, you are building your robot and control stack from the ground up. Our role is to ensure the operating system layer delivers everything required for your unique environment: low jitter, security, and functional safety.”
A new frontier for manufacturing transformation
Foundational software is not always front of mind for decision-makers charting the path of digital transformation in manufacturing. Yet its influence extends to almost every dimension of performance: speed, precision, safety, scalability, and even compliance.
Robots that move safely and efficiently in human spaces react precisely to sensor input and adapt to changing workloads with AI-enhanced autonomy. All of these outcomes depend on software foundations that are fit for purpose. While real-time responsiveness might seem like a small technical detail, it is measured in milliseconds, which defines competitive advantage.
“The blink of an eye takes about 100 milliseconds,” Leung notes. “That is already too slow for some robotic applications. In environments like robotic surgery, you would feel the delay. As robots become more intelligent and autonomous in manufacturing, the need for near-microsecond responsiveness only becomes more critical. It is not about the robot learning independently but about the system responding fast enough for AI to be useful.”
The industry is heading toward a future where intelligent automation is expected rather than exceptional. In that future, foundational software is not just the quiet layer underneath, it is the critical enabler of safe, precise and adaptable manufacturing. Those who invest early in this capability will be best positioned to lead it.
