A spellbinding tale of how a university group created a fully autonomous warehouse solution for real-time inventory accuracy
A warehouse may not seem like the most titillating environment, but its big open space and four walls are very much the life force of the manufacturing industry—from logistics to supply chain operations, an organized (or not so organized) warehouse can make or break your operations.
How so? If you’re unable to keep track of the goods in your warehouse, you’ll most likely have inventory inaccuracies. These inaccuracies can lead to stockouts, misplaced goods, and incorrect fulfillment, which costs businesses over $1.77 trillion annually in operational disruptions.
Poor space utilization also reduces warehouse efficiency by up to 15%, creating bottle necks and slowing productivity. Manual inventory checks are time consuming and labor intensive, consuming up to 55% of operational expenses while remaining prone to errors.
This is why students from University of Pennsylvania entered the SICK $10K Challenge, a contest for universities across the nation to put their innovation skills to the test by finding a real-world problem and creating a solution for it with LiDAR technology.
“The industry demands a smarter, autonomous, and scalable solution that eliminates inefficiencies, optimizes space, enhances safety, and improves inventory visibility,” said team member Kevin Paulose.
Providing warehouse professionals with real-time inventory accuracy drove the team to create Nimbus.
What is Nimbus?
Navigation and Inventory Monitoring Bot for Utility and Surveillance (Nimbus) is a fully autonomous warehouse solution designed to provide real-time inventory tracking, optimize space usage, enhance safety, and flag anomalies before they disrupt operations.
“Inspired by Harry Potter’s Nimbus 3000, our Nimbus robot brings the same speed, agility, and precision to warehouse inventory monitoring, ensuring a warehouse is always in motion, always monitored, and always optimized,” Paulose said.
The workings of an autonomous warehouse bot
It may not require the swish of a wand, but replacing manual hours-long searches for inventory with an autonomous robot sure feels like magic. Nimbus utilizes SICK’s multiScan100 sensor, stereo cameras, and multi-model perception to navigate its surroundings and detect anomalies. These anomalies could be anything from protruding boxes to obstacles on the bot’s path, contributing to a safer working environment.
And better yet, warehouse professionals can visualize their environment with NimbusView. No need to divine the future from tea leaves or a murky crystal ball when you have an AI-powered digital twin and inventory monitoring software.
Operators can see every product's exact location, track inventory flow, and optimize space utilization. Misplaced stocks? No problem. NimbusView automatically flags incorrect placements, reducing errors and improving accuracy. When a misplaced item feels like an impossible-to-find golden snitch, NimbusView is up to the task.
Warehouse safety: No Skele-Gro needed
Beyond misplaced stock and supply chain bottlenecks, safety hazards are also a big concern in warehouse and distribution center environments. OSHA reports that the warehousing industry experiences 4.8 injuries per 100 full-time workers, and 36% of nearly 100,000 forklift accidents result in serious injury or death.
With Nimbus's ability to detect anomalies, like protruding boxes or obstacles on its path, it quite literally clears the way for a safer environment.
The need for Nimbus
While other methods are in place to track inventory—like barcode scanning, manual cycle counting, and RFID systems—they often fail to provide accurate insights in real time. Team member Tejendra Patel explained the warehouse automation market is experiencing shortages and increasing e-commerce demands, which means a more reliable process is necessary.
Nimbus also provides this much-needed process upgrade at a reasonable cost and with feasible deployment.
“Nimbus operates on a robot-as-a-service model, replacing capital expenditures with operational expenses, [which] safeguards businesses investments, unlike expensive fixed automation. It’s flexible, scalable, and instantly deployable, making AI-driven warehouse intelligence accessible to midsize- and enterprise-level warehouses alike,” Paulose said.
The University of Pennsylvania won third place in the SICK $10K Challenge with their strong business case and inventive use of LiDAR. While they may not have received a letter from Hogwarts, their admission into the contest spurred a new level of innovation—the real sorcery of it all.
