Pakistani robotics data company World Context organized its first robotics hackathon at the University of California, Berkeley, uniting 13 teams to foster stronger artificial intelligence and technology connections between Pakistan and the United States.
The event was conducted in collaboration with Blockchain at Berkeley and supported by sponsors including Google Gemini, ElevenLabs, The Hackathon Company, Lovable, Praxis Robotics, and UC Berkeley Startup. Participants utilized datasets provided by World Context to develop projects spanning hardware, software, and visualization categories.
Sortbot, a project aimed at creating robots capable of sorting various objects, secured the hardware track victory. The Reconstructors team, which focused on reconstructing World Context data into physical environments, won both the software track and the visualization track with a simulation they developed.
A total of $4,500 (Rs1.25 million) in cash prizes were awarded during the competition.
The hackathon underscored the increasing significance of human demonstration data in robotics. By capturing how individuals perform tasks in real-world settings, this data provides developers with practical examples of physical task execution, aiding in the creation and training of systems designed for operation in physical environments.
World Context gathers human demonstration data from workers in real-world contexts and markets these datasets to robotics companies, viewing such data as a vital component of the evolving robotics ecosystem.
Founder and CEO Sherdil Ali Khan expressed surprise at the high quality of submissions and announced plans to collaborate with universities to host similar events in the future. He noted, “The talent showcased at Berkeley reflects the advanced state of the industry in the U.S., presenting great opportunities for learning and partnership with American institutions and companies. We aim to bring these insights back to Pakistan to enhance its technological landscape.”
Khan also highlighted the creativity and technical skills demonstrated by participants, who applied World Context datasets in innovative ways beyond the company’s expectations.
Looking ahead, World Context intends to pursue closer cooperation with other leading universities. The company emphasized that large-scale visual datasets will remain crucial for robotics development and that providing students access to such data offers valuable early exposure to the technology.
UC Berkeley was chosen for this inaugural hackathon due to its strong robotics and engineering programs and the chance to engage directly with its students and researchers.
The event also provided Khan with opportunities to connect with U.S.-based robotics and computer vision engineers and companies. He held discussions with several data and robotics firms about potential collaborations in areas such as Vision-Language-Action (VLA) systems, egocentric and exocentric data collection, and hardware deployment.
The hackathon demonstrated how human demonstration data can bridge the divide between robotics software and hardware, a synergy World Context expects to grow increasingly important as the industry advances.
Leave an opinion