VIVAERO

Student proof of concept pairs virtual reality with warehouse drones

University of Cincinnati engineering students have developed a proof-of-concept system that combines autonomous drones, laser-based navigation and a photorealistic virtual-reality model for warehouse inventory tasks. Working with Valvoline, doctoral students Deepak Antony David and Luke Busse designed the system so a worker can explore a virtual representation of a warehouse and dispatch a drone to an aisle or section to look for items or inspect potentially damaged pallets. The students planned to present the project at the Modeling, Estimation and Control Conference on October 25 in Phoenix, Arizona.

At the center of the prototype is a warehouse digital twin: a photorealistic model that provides an interactive visual representation of the facility. The interface runs on Unreal Engine, software also used in games such as Fortnite and Batman: Arkham City, and displays a live, third-person video of the virtual drone as the physical aircraft moves through the corresponding space. Busse said the high-fidelity presentation resembles a video game.

Its stated function is to give warehouse workers an intuitive way to inspect the modeled environment and direct an autonomous drone to a selected location without manually steering the aircraft. The prototype drone navigates indoors using light detection and ranging, or LIDAR, rather than satellite positioning. It emits pulses of laser light to detect obstacles, which the source describes as supporting obstacle avoidance within centimeters in warehouse aisles.

University of Cincinnati student engineers designed an inventory management system for warehouses that combines autonomous drones and photorealistic virtual reality. The drone uses LIDAR to navigate the tight confines of warehouse aisles(image above). Credit: Michael Miller

The system was refined in UC’s high-bay Digital Futures laboratory, where the students built a mock warehouse inside a flight cage surrounded by protective safety webbing. Rows of pallets filled with empty boxes represented a storage environment. During the demonstration, the drone used LIDAR to capture the contours of the space, and a colorful, real-time three-dimensional rendering appeared on a laptop while the physical drone moved around the mock warehouse.

Inventory accuracy is the operational problem identified by the team. Busse said the students found that 65% of warehouses in North America still use manual inventory, with workers potentially categorizing products incorrectly, putting items in the wrong place or spending time correcting mistakes. The drone system is also intended to alert workers to leaks from containers that could damage entire pallets.

These capabilities remain part of a proof-of-concept collaboration with Valvoline through UC’s Industry 4.0/5.0 Institute; the supplied report does not describe a commercial deployment or quantify improvements in inventory accuracy, inspection speed or leak detection. The team also expects interest in automated inventory systems to grow as retailers and other industries build more warehouses. Interact Analysis, as cited in the university report, estimates that more than 4,000 new warehouses will be built in the United States over the next four years.

That estimate describes future construction rather than evidence about the prototype’s performance.

How the system works:

• LiDAR navigation: Because indoor warehouses block satellite signals (GPS), the drones use Light Detection and Ranging (LiDAR) laser pulses to map aisles, fly within centimeters of accuracy, and avoid obstacles.

• Digital twins & VR: The drones operate inside a photorealistic 3D digital twin of the facility. Workers use VR interfaces or computer views to inspect high shelves, monitor stock, and interact with the space virtually.

• Leak and hazard detection: Beyond counting stock, the drones scan for container leaks or damaged pallets to prevent spills from ruining entire stacks.

Key benefits:

• Speed and efficiency: Automated drone scanning can be up to 15 times faster than manual counts using lifts and ladders.

• High accuracy: Inventory tracking accuracy regularly exceeds 99%, reducing discrepancies and overselling.

• Safety: Workers are removed from dangerous high-bay climbing tasks and hazardous zones.

• Continuous operations: Counts run in the background with little to no facility downtime or disruption to daily forklift and worker traffic

University of Cincinnati