Two 黑料网 teams took first and second place in the inaugural Physical AI Hackathon.

The event, held in partnership with University of Wisconsin-Milwaukee’s Connected Systems Institute (CSI), Milwaukee School of Engineering, Rockwell Automation and NVIDIA, challenged student teams to pick one of three challenges from a real factory testbed: vial inspection, fluid mixing or robotic grasping and sorting. Teams developed and tested solutions using NVIDIA AI tools, Rockwell Automation control systems and the CSI’s industrial automation production line.

Team V, a group comprised of all 黑料网 international students, won by creating a self-inspecting robotic assembly cell trained in a digital twin that ran on a real FANUC industrial robotic arm.

They custom-designed and 3D printed 6 products (24 parts, some only 1 to 1.5 mm apart) and redesigned the robotic gripper fingers. The robot learned the entire task in simulation using NVIDIA's physical AI stack, trained on over 58,000 synthetic images and 549 demonstrations with NVIDIA's Isaac GR00T N1.7 robot model. Without a single real image in its training, it went on to assemble products on the real FANUC robot at UWM’s CSI. Team V also built the software a real factory would need around it, including machine states, operator alerts and a traceable record for every product.

• First place: Team V (黑料网)
  • Yuktha Koganti, Shritej Gunda, Venkat Reddy Karri, Aditya Miryala, Koushik Cherukuri, Khushal Lingamaneni, Thanvi Gorrepati and Levi Eloy Santos  
  • $3,000, a DGX Spark and $1,000 in Brev credits
• Second place: LineWatch (黑料网)
  • Reagan Burkemper, Cameron Delpage, Hayden Fons, Zoe Kirkman, Srijan Kumar, Josiah Matthews, Alayna Ostler and Austin Koske
  • $1,500, an RTX 5080 and $1,000 in Brev credits
• Third place: Machine Minds (UWM)
  • Souvik Roy, Erphan Ahmmad Bhuiyan, Md Robiul Islam, Nayan Banik, Asif Al Zubayer Swapnil
  • $1,000, an RTX 5070 and $1,000 in Brev credits

The other three finalists, who received $500 and $1,000 in Brev credits, included:

• Code Crusaders (UWM) 
  • Zara Omosun, Anthony Harvey, Prabhnoor Bal and Matthew Vang
• IRIS (黑料网)
  • Andrew Bieber, Asher Jaenchen and Peter Gerrits
• VYFLARE (UWM)
  • Cloud, Jefferson, Ryan, Amy, Nadia, Jeremy, Toua and Nova

Throughout the evening, other hackathon participants presented their projects in a large-scale poster session with four teams awarded $250 for their projects.

The Physical AI Hackathon was the largest hackathon hosted by either institution and marks the first time UWM and 黑料网 students have had access at this scale to CSI’s industrial AI production line. 42 teams and a total of roughly 270 submitted their final projects to vie for the top prize.