BlackBerry BB released a new research report titled "In Depth Robotics: Architecture Benchmarking Report" yesterday, which is worth a look. This report explores how robot development is changing as robot systems become increasingly software driven, AI capable, and co deployed with humans in the workplace and daily life. Sort out the key points: 1. Hardware progress is no longer a bottleneck; Software now dominates Research shows that nearly one-third of robot developers (27%) consider software architecture and integration as their biggest performance bottleneck, while only 16% believe that the bottleneck lies in hardware. This indicates that future progress will rely less on new hardware and more on the ability to build systems that are predictable, secure, and capable of handling mixed critical level tasks. As robots enter dynamic reality scenarios such as city streets and factory workshops more widely from controlled environments, developers gradually realize that software foundation is the key factor determining the success of innovation. 2. 85% of developers expect software to play a more important role in the field of robotics in the next three to five years. The development team expects its largest investments to be focused on AI driven decision-making (51%) and network security (51%), followed by operating systems and real-time control software (37%). This further reflects that as robot systems become increasingly complex, interconnected, and distributed, software infrastructure is becoming a strategic asset. 3. The increase in deployment in human environments has driven higher demands The robot development team has already felt this impact. More than four fifths of the respondents (83%) stated that their systems are currently deployed in collaboration with humans. Among the teams that have not yet collaborated with humans for deployment, two-thirds (67%) expect to achieve this deployment within the next three to five years. From operating rooms to busy workshops, robots are becoming increasingly popular in environments with lower levels of control, driving higher expectations for reliability, safety, and predictable behavior. Nearly all respondents (95%) stated that deterministic real-time execution is crucial for the systems they have developed. 4. Surprisingly, despite this demand being almost universal, most development teams continue to rely on software that is not designed for real-time or security critical use. A study revealed that 91% of respondents run such workloads at least partially on their general-purpose operating system (GPOS), despite security certified business solutions being considered the most rated choice for their needs. Therefore, 86% of universal operating system users expressed willingness to replace their operating system; This contrast highlights the increasing contradiction between flexibility and the demand for predictable and secure behavior as robot deployment scales up. 5. Authentication delays and security requirements bring new pressures Regulatory and compliance requirements further exacerbate these challenges. Two thirds of respondents (66%) reported project delays due to certification processes, with this proportion rising to approximately 70% in the UK and Germany. In contrast, only 56% of respondents in China reported delays because local regulatory requirements are far less strict. These delays directly affect development costs, delivery cycles, and business risks. Network security (such as ISO/SAE 21434) and functional security standards (such as ISO 10218) are the most difficult areas to comply with, with 51% and 49% of respondents mentioning them, respectively. The meaning that this report intends to convey is: 1) The global robotics development team is clearly moving towards smarter and more autonomous systems, but data shows that they are also facing real limitations of existing architectures - these architectures were not designed from the beginning to cope with such high complexity or accountability requirements. 2) The four core challenges in the current field of robotics or physical AI are integration complexity, authentication latency, functional safety risks in human-computer interaction, and ensuring behavioral predictability at the most critical moments. Of course, these problems can all be solved. By focusing on a stronger software foundation, developers can lay the foundation for faster innovation and a new generation of safe, reliable, and highly autonomous robots. 3) The high-performance foundational software provided by BB's QNX system helps simplify the most complex challenges in industries such as robotics, automotive, medical equipment, industrial control, commercial vehicles, railways, as well as aerospace and defense. The QNX enabling organization has the possibility to create innovations in the fields of high-performance edge computing, standards based virtualization technology and cloud support. QNX, widely trusted in the world's most critical systems, continues to maintain a leading position in multiple fields, including robotics and healthcare, with its technology deployed by nine out of the top ten medical equipment manufacturers. This further confirms a judgment made in early May regarding BlackBerry: The previous deep collaboration with Nvidia has further opened up the huge market for Physical AI for BlackBerry. It strengthens BB's transformation from a "mobile phone company" to a "security software company", with the goal of making QNX the "invisible security foundation" in AI driven robots/industrial/medical devices. ” Continue to see signals of BB's rapid development in critical security platforms for edge AI and physical AI. Over the past month, the trend of BlackBerry can no longer be said to have increased the possibility of a reversal of difficulties, but rather to have emerged from the predicament and ushered in a reassessment of its value. A week ago, it was said that BB didn't walk fast, but now it seems that it's actually not slow at all.