Dr. Gulam Dastagir Khan
Assistant Professor
Department of Electrical and Computer Engineering, College of Engineering
Most of us use a lift several times a day without giving it much thought. We press a button, wait for the doors to open, step inside, select a floor, and continue our journey. For a robot, however, this seemingly simple activity becomes an extraordinary engineering challenge. How does it recognise the lift? How does it know where to stand? How can it press the correct button with precision? How does it determine when the doors have opened? And once it reaches another floor, how does it continue its work without human assistance? These were some of the questions that motivated our recent research at SQU, where my research team has been exploring how AI, robotics, and autonomous systems can support smarter and more efficient building management.
Looking beyond industrial robots
When people think of robots, they often imagine machines working inside factories, assembling products with speed and precision. Service robots face a very different challenge. They must operate in environments designed for people rather than machines. Hospitals, universities, airports, hotels, shopping centres, and government buildings are dynamic environments where robots must navigate corridors, avoid obstacles, interact safely with existing infrastructure, and make intelligent decisions in real time. Developing robots capable of working in these environments requires much more than mobility. They must be able to understand their surroundings, interpret visual information, and interact with everyday objects that humans rarely notice.
The challenge that changed the project
During the early stages of the research, one practical limitation quickly became apparent.
Most autonomous robots can work efficiently on a single floor, but very few can move independently between floors in conventional buildings. Existing solutions often rely on specially designed "smart" lifts connected to digital control systems, making them unsuitable for the majority of buildings currently in use. Instead of asking how buildings could be modified for robots, we asked a different question:
Can we teach robots to adapt to existing buildings instead?
Answering this question became one of the defining objectives of the project. Using onboard cameras, computer vision, AI, autonomous navigation, and a lightweight robotic manipulator, we developed a robotic system capable of recognising and operating conventional passenger lifts without requiring expensive modifications to the building. Although pressing a lift button appears trivial, achieving this reliably required the integration of perception, navigation, manipulation, and decision-making into a single autonomous system.
Beyond mobility
Using a lift was only one part of a much broader vision. The ultimate objective was to develop a service robot capable of assisting engineers with routine building inspections. The robot uses AI-powered vision to recognise fire extinguishers, lighting fixtures, and visible building defects such as cracks, moisture damage, peeling paint, and stains. Instead of simply capturing photographs, it identifies the location of these observations and automatically generates inspection reports that can assist maintenance teams in planning repairs and preventive maintenance. By performing repetitive inspection tasks autonomously, robots can allow engineers to spend more time analysing problems and making informed decisions rather than collecting routine information.
From the laboratory to a real building
Research in robotics cannot stop at simulation. Algorithms that perform well inside a computer often face entirely different challenges when deployed in real environments. For this reason, our research focused heavily on experimental validation.
The complete robotic platform was designed, assembled, integrated, and tested in a real multi-storey academic building. During extensive field trials, the robot successfully navigated corridors, located inspection targets, travelled between floors using a conventional passenger lift, and completed inspection tasks without human intervention. Watching the robot successfully call a lift, enter the cabin, select the correct floor, and continue its mission autonomously was a particularly rewarding moment for the research team. It represented the culmination of months of design, software development, integration, testing, and refinement. For researchers, these moments provide confirmation that an idea has moved beyond theory and become a practical engineering solution.
Learning through research
One of the most rewarding aspects of research is the opportunity it provides for students to tackle genuine engineering problems. Projects such as these expose students to robotics, AI, computer vision, embedded systems, software engineering, and experimental validation within a single multidisciplinary project. Rather than learning technologies independently, students experience how multiple engineering disciplines come together to solve real-world challenges. This practical experience prepares them to contribute to emerging industries where AI and robotics are becoming increasingly important.
Supporting Oman Vision 2040
The research also reflects the broader ambitions of Oman Vision 2040, which emphasises innovation, digital transformation, scientific research, and the development of a knowledge-based economy. Universities play a central role in achieving these goals by transforming research into practical technologies that address national priorities and strengthen local expertise in advanced engineering fields. Research should not end with publications. Its greatest value lies in creating knowledge that benefits society, supports industry, and inspires the next generation of engineers and innovators.
Looking ahead
The intelligent robots being developed today are only the beginning. Future generations of service robots will become more capable, more autonomous, and more closely integrated with smart buildings and digital infrastructure. They will assist engineers by performing repetitive tasks, collecting reliable information, and improving the efficiency of facility management across hospitals, universities, airports, and public buildings. At SQU, we are proud to contribute to this journey through research that combines scientific innovation with practical impact.
After all, teaching a robot to use a lift is not really about lifts.
It is about creating intelligent machines that can safely work alongside people, solve real problems, and help build a smarter future for Oman.