Seungyoung Park, Senior Research Engineer
HL Robotics, a developer of indoor and outdoor parking robots, is advancing its automation capabilities by building an integrated operations system on Amazon Web Services (AWS). At the AWS Summit Seoul 2025, HL Robotics unveiled its strategy to enhance solution maturity through four core components: hybrid cloud, robot DevOps, system monitoring, and simulation.
“We believe real-time control is the essence of robotics,” said Seungyoung Park, Senior Research Engineer at HL Robotics, during an interview with Bloter at the summit held in Gangnam, Seoul on May 14. “Control should take place on a private network with operations and management handled through the public cloud.”
As HL Robotics continues to refine its indoor parking robot “PARKIE” and builds upon its 2023 acquisition of Stanley Robotics—the world’s first commercial outdoor autonomous parking company—the company is evolving toward a unified indoor-outdoor robotic parking solution. Park leads the overall robot control system at HL Robotics, a robotics subsidiary of the HL Group, which includes HL Mando and HL D&I Halla.
Hybrid Cloud, Essential for Robot Operations
To achieve both stability in control and flexibility in operations, HL Robotics has adopted a dual-structure approach: robot control runs on a private network, while operational commands and monitoring are handled via the public cloud. “Because low-latency response is critical for parking robots, real-time control happens on a private network,” Park explained. “Meanwhile, monitoring, updates, and anomaly notifications are managed through the cloud.”
This architecture addresses both the complexity and security demands of real-world parking environments. “Every parking site has different requirements,” Park added. “While real-time control requires minimal latency, video surveillance and operational systems must be centrally managed.”
Parking robots are not just about moving cars. They are complex systems encompassing control, autonomous navigation, and user recognition—each relying on different software services. As such, higher security and better complexity management are needed compared to typical single-network structures. HL Robotics meets these demands with a hybrid cloud-plus-private network model.
Even Robots Need DevOps
Seungyoung Park, Senior Research Engineer
Park also emphasized the importance of scalability and efficiency in robot software deployment. “Robotic operations require various software components to run simultaneously,” he said. “We need a deployment and update framework tailored for each site—what we call ‘Robot DevOps.’”
Real-world deployments face practical constraints, such as mechanical parking regulations or requirements to separate human and robot movement. “There are times when a parking spot is temporarily closed or a robot’s path needs to be updated on the fly,” Park said. “And where robots and people interact, strict safety protocols must be enforced.”
These changing conditions must be reflected in software in real time, making flexible deployment environments essential. “Even when development happens locally, the software needs to be distributed and updated properly at the required sites,” Park added. “We found AWS offered many tools that made this possible.”
Real-Time 3D Monitoring and Simulation, All in the Cloud
Enhancing the robot monitoring system is another focus. HL Robotics has developed a 3D visualization interface that allows operators to view the status of robots and vehicles in real time. “In unpredictable environments like parking lots—where people may enter unexpectedly or foreign objects appear—video monitoring and alerts are critical,” said Park. “We’ve built a system that receives sensor and CCTV data via the cloud and delivers real-time alerts to operators.”
HL Robotics is also using the cloud to run simulations for pre-deployment validation and customer proposals. “When proposing a solution, we need to quantify improvements in wait times, space utilization, and robot routing,” Park explained. “We can simulate multiple scenarios in parallel and automatically visualize results, which strengthens the credibility of our proposals.” These simulation capabilities now support not only technical design, but also customer-specific service planning and investment decision-making.
Looking ahead, HL Robotics is developing systems for predictive maintenance and lifecycle analysis. By collecting sensor data and operation logs through the cloud infrastructure, the company is building models to detect anomalies and forecast maintenance schedules. “The goal is to secure high-quality data and ultimately automate the operation process,” Park said.
In closing, Park added: “We’re no longer in an age where we lack technology. There are too many options now—so what matters is choosing the right one. Working with AWS helped us validate and apply what we needed quickly, which is how we continue to refine and complete our solutions.”
Seungyoung Park, Senior Research Engineer
HL Robotics, a developer of indoor and outdoor parking robots, is advancing its automation capabilities by building an integrated operations system on Amazon Web Services (AWS). At the AWS Summit Seoul 2025, HL Robotics unveiled its strategy to enhance solution maturity through four core components: hybrid cloud, robot DevOps, system monitoring, and simulation.
“We believe real-time control is the essence of robotics,” said Seungyoung Park, Senior Research Engineer at HL Robotics, during an interview with Bloter at the summit held in Gangnam, Seoul on May 14. “Control should take place on a private network with operations and management handled through the public cloud.”
As HL Robotics continues to refine its indoor parking robot “PARKIE” and builds upon its 2023 acquisition of Stanley Robotics—the world’s first commercial outdoor autonomous parking company—the company is evolving toward a unified indoor-outdoor robotic parking solution. Park leads the overall robot control system at HL Robotics, a robotics subsidiary of the HL Group, which includes HL Mando and HL D&I Halla.
Hybrid Cloud, Essential for Robot Operations
To achieve both stability in control and flexibility in operations, HL Robotics has adopted a dual-structure approach: robot control runs on a private network, while operational commands and monitoring are handled via the public cloud. “Because low-latency response is critical for parking robots, real-time control happens on a private network,” Park explained. “Meanwhile, monitoring, updates, and anomaly notifications are managed through the cloud.”
This architecture addresses both the complexity and security demands of real-world parking environments. “Every parking site has different requirements,” Park added. “While real-time control requires minimal latency, video surveillance and operational systems must be centrally managed.”
Parking robots are not just about moving cars. They are complex systems encompassing control, autonomous navigation, and user recognition—each relying on different software services. As such, higher security and better complexity management are needed compared to typical single-network structures. HL Robotics meets these demands with a hybrid cloud-plus-private network model.
Even Robots Need DevOps
Seungyoung Park, Senior Research Engineer
Park also emphasized the importance of scalability and efficiency in robot software deployment. “Robotic operations require various software components to run simultaneously,” he said. “We need a deployment and update framework tailored for each site—what we call ‘Robot DevOps.’”
Real-world deployments face practical constraints, such as mechanical parking regulations or requirements to separate human and robot movement. “There are times when a parking spot is temporarily closed or a robot’s path needs to be updated on the fly,” Park said. “And where robots and people interact, strict safety protocols must be enforced.”
These changing conditions must be reflected in software in real time, making flexible deployment environments essential. “Even when development happens locally, the software needs to be distributed and updated properly at the required sites,” Park added. “We found AWS offered many tools that made this possible.”
Real-Time 3D Monitoring and Simulation, All in the Cloud
Enhancing the robot monitoring system is another focus. HL Robotics has developed a 3D visualization interface that allows operators to view the status of robots and vehicles in real time. “In unpredictable environments like parking lots—where people may enter unexpectedly or foreign objects appear—video monitoring and alerts are critical,” said Park. “We’ve built a system that receives sensor and CCTV data via the cloud and delivers real-time alerts to operators.”
HL Robotics is also using the cloud to run simulations for pre-deployment validation and customer proposals. “When proposing a solution, we need to quantify improvements in wait times, space utilization, and robot routing,” Park explained. “We can simulate multiple scenarios in parallel and automatically visualize results, which strengthens the credibility of our proposals.” These simulation capabilities now support not only technical design, but also customer-specific service planning and investment decision-making.
Looking ahead, HL Robotics is developing systems for predictive maintenance and lifecycle analysis. By collecting sensor data and operation logs through the cloud infrastructure, the company is building models to detect anomalies and forecast maintenance schedules. “The goal is to secure high-quality data and ultimately automate the operation process,” Park said.
In closing, Park added: “We’re no longer in an age where we lack technology. There are too many options now—so what matters is choosing the right one. Working with AWS helped us validate and apply what we needed quickly, which is how we continue to refine and complete our solutions.”