Why hasn’t Augmented Reality (AR) had a greater uptake in the manufacturing market? Maybe it had not found the best use. Data Time Space believes they may have found the killer app for AR: speeding continuous improvement (CI) and freeing CI experts to do the analysis, not the data collection. This young company enables manufacturers to use AR in new ways on mobile devices to collect real-time data over time in 3D space.
Spatial-Context Data Collection for CI and More
Data Time Space has developed a collaborative AI-powered spatial computing platform that collects data within a 3D video scan from a mobile device. The platform’s data model and AI capture, store, normalize, and put in context spatial data. The platform is inherently designed to connect to the Internet of Things (IoT) and track human movement. All computing in the platform is 3D space-related.
The company is initially aiming it at measuring ergonomics and safety, conducting time-and-motion studies, and gauging how people and processes are performing. The common thread for these applications is that they need data with spatial context to succeed. Seeing how a worker performs in 3D space is one thing. Capturing all of the data in context, based on an actual video scan of any part of a factory, and normalizing it is what Data Time Space is offering. The resulting type of dataset is foundational for all of these issues. The accurate, unbiased, contextualized data set has been complicated for most CI teams to assemble until now.
Mobile and 3D
Many AR applications’ hardware is based either on headgear or fixed cameras. Headgear can be uncomfortable, and cameras are inflexible and intrusive if always on in a work area. Data Time Space uses tablets or mobile phones. They point out that mobile devices can now be located within a few centimeters, with accurate LiDAR depth perception for precise signal localization in the camera. Signals are object locations or movement deltas. This spatial accuracy and the flexibility of using it where and only when needed could significantly improve AR’s spatial-context data’s accuracy, comfort, and privacy.
Rather than 2D maps and analysis tools, everything starts and stays in realistic 3D. End-to-end 3D eliminates transformations and the delays and possible errors they introduce. It also facilitates understanding and streamlines data gathering and analysis. The model trains on pointing the system at the user or process in question and pushing a button to record the data in 3D in real-time, time series views.
AI’s Role
While the platform is the major differentiator, one of its capabilities that may raise questions is spatial AI. AI helps with input and output to the spatial data model. AI collects spatial and temporal data, understanding what is in front of the camera and localizing it in 3D space. AI also uses and analyzes data in the model. A large language model (LLM) enables queries and processes data stored in the model.
Supporting Experts
It is designed to make the experts’ lives easier. If these professionals have the data collected and analyzed in real time, they can focus on how to best solve the problem. Data Time Space points out that the AI in their platform helps to understand the world and probe the complex data set, but it is human-in-the-loop and data model-centric. So, it has a strong model with logic in the platform and a streamlined, researched human interface.
Another challenge with CI data collection and analysis is bias. Using the Data Time Space system quickly alleviates any concern about that, as everyone sees exactly what happened in the recorded data. The demo is compelling, showing it tracking movement for ergonomics and trace or spaghetti maps for movement and waste reduction.
Key Applications
Data Time Space has also created AI-based applications to run on the platform. The main applications are factory analytics for continuous improvement and remote assistance. Factory analytics could compare differences between shifts, workers, or facilities. The focus could be on eliminating wasted movement, ergonomics, or safety. Remote support enables experts to stay in one location and support production teams worldwide, as they see exactly what is happening. The visual nature of the system and its data capture means it can also support employee training and education.
Data Time Space says it has found a way to scale applications that conform to its spatial data model flatly. It expects to build more applications or tools that help users. The applications can play together, enabling a remote expert to perform CI analysis without being physically present, for example.
New Yet Proven and Honed
While Data Time Space is a young company, the founders have been focused on building and enhancing this technology for 12 years. Before this, they worked at PTC, building and honing the Vuforia AR/VR platform. Data Time Space now owns the IP behind this spatial platform, while PTC owns a share of the company. This is a complete spatial data collection and analysis platform, with AI incorporated to understand the spatial data better. The user interface also includes AI. With their deep background, they have quickly begun to offer proofs of concept (PoC) that hold promise.
Looking Forward
CI data collection may well be the killer app for AR. Data Time Space’s next steps will be helping to prepare the data sets for uses beyond the CI team. Thank you, Valentin Heun, for taking the time to explain your approach and technology to us. We look forward to reconnecting and learning more as you make progress in the market.