DreamVu
"We capture the physical world in 3D"
Philadelphia firm founded 2017 that captures real venues with its own single-shot 360-degree stereo camera and sells world-model and robot training data.
DreamVu, Inc. is an applied research company that captures real places in 3D with its own 360-degree stereo camera. It sells the output as world-model data, robot training data and simulation environments for physical AI. It is headquartered in Philadelphia, Pennsylvania, with a west-coast office in Palo Alto, California and an R&D center in Hyderabad, India. The company grew out of a computer vision research group at IIIT Hyderabad with about ten years of published work in computational imaging, depth estimation and 3D reconstruction.
The company was founded in 2017, took part in the AVISHKAR deep-tech accelerator at the IIIT-Hyderabad Foundation, and received strategic funding from SRI Capital in December 2018. In January 2021 it launched the PAL and PAL Mini omnidirectional 3D vision systems for mobile robots. Each uses a single CMOS sensor with no moving parts and outputs color video with depth in one stream. PAL offered up to 9 MP with a 360 x 110 degree field of view and 10 m depth range for USD 1,499. PAL Mini offered up to 4 MP, 360 x 80 degrees and 5 m depth range for USD 999. Both had USB interfaces and an SDK supporting ROS and OpenCV. DreamVu was named an RBR50 company by The Robot Report in 2021. In January 2026 the business relaunched around physical-AI data.
Current offerings are built on the Alia camera system, a patented optical design that captures 360 degrees in stereo in a single shot and produces aligned color and depth through the same optics. DreamVu sells data rather than cameras. World Model Data delivers 16K equirectangular color imagery with metric depth at every pixel, plus depth maps and point clouds from real locations; a sample corpus, Liberty3D, covers landmarks in Philadelphia, New Jersey and Washington, DC. Robot Training Data combines an exo-view 360-degree stereo panorama with depth, a synchronized head-mounted ego camera and an optional wrist camera. Annotations include capability-probed Q&A pairs for VLMs, trajectory and 6-DoF pose tracking for navigation, and hand pose, contact events, action boundaries and task breakdowns for loco-manipulation. Data ships in LeRobot (RLDS), Open X-Embodiment or custom schemas, and an open PRISM-100K subset is on Hugging Face. Target settings include grocery, pharmacy, hardware retail, fashion, restaurant kitchens, manufacturing, warehousing, clinical and home care. DreamVu also reconstructs real venues as simulation environments and publishes research on video world models, multi-view datasets and VLA models.
Leadership includes Sashi Reddi (Chairman and CEO; founder of SRI Capital), Rajat Aggarwal (Chief Technology Officer), Parikshit Sakurikar (VP Imaging and AI), Anoop Namboodiri (co-founder, scientific advisor, professor at IIIT Hyderabad) and Tony Bartolucci (VP Engineering). Earlier press names Rajat Aggarwal and Rohan Bhatial as founders. Investors include SRI Capital, Ben Franklin Technology Partners of Southeastern Pennsylvania, Broad Street Angels and Chiratae Ventures. The company has raised about USD 4.56 million in total, including a USD 2 million round on April 20, 2026, and has about 28 employees.
Primary type & automation activities this supplier delivers:
More related news
- 1
- 2
- 3
-
4
CASIA's Peiyan Li Brings Two VLA Papers to IROS 2026 RobotToday · Oct 08, 2026
- 5
- 6
- 7
- 8
- 9
- 10
- 11
-
12
IROS 2026 Best Paper Finalists: South Korea Leads RobotToday · Oct 07, 2026
- 13
-
14
IMTS 2026 Preview: Record Orders and the Automation Story RobotToday · Sep 14, 2026
-
15
Software, Compute and Interaction at WRC 2026 RobotToday · Aug 31, 2026
-
16
WRC 2026 Exhibitor Map: 248 Robotics Companies Mapped RobotToday · Aug 26, 2026
-
17
Inside Saga Robotics: UVC Light vs. Crop Disease RobotToday · Aug 14, 2026
- 18
-
19
WAIC Review 2/4: World Models vs. VLA: Robotics' Next 'ChatGPT Moment' RobotToday · Aug 03, 2026
- 20
Contact DreamVu
WEBSITE
https://www.dreamvu.ai/HEADQUARTERS
United States
Company Facts
Founded
2017
Primary Role
Software/Algorithm
Company Size
employees 10-50
Primary Region
North America
Annual Sales
-
Funding Stage
Growth / Late-Stage VC
Funding Total
$ 4,560,000
Keywords
- DreamVu
- Inc.
- DreamVu Inc.
- DreamVu AI
- omnidirectional stereo camera
- 360-degree depth
- world model data
- robot training data
- egocentric and exocentric video
- VLA datasets
- simulation environments
- LeRobot format
- Software/Algorithm
- Component Supplier
- Cloud & Data
- Perception & Vision
- Retail & Hospitality
- Manufacturing
- Logistics & Supply Chain
- Healthcare & Senior Care
