PIWIMS: Physics Informed Warehouse Inventory Monitory via Synthetic Data Generation

Abstract

State-of-the-art camera-based deep learning methods for inventory monitoring tend to fail to generalize across different domains due to the high variance of scene settings. Large amounts of human labor are required to label and parameterize the models, making a real-world deployment impractical. In a third-party warehouse setting, supervised learning approaches are either too costly and/or inaccurate to deploy due to the need for human labor to address the diverse set of environmental factors (i.e, lighting conditions, product motion, deployment limitations).We introduce PIWIMS, a realistic synthetic dataset generation technique that combines the physical constraints of the scene in real-world deployments, drastically reducing the need for human labeling. In contrast to other generative techniques, where the generative parameters are learned from a large sample of available data, this compositive approach defines the parameters based on physical characteristics of the particular task, which requires minimal human annotation. We demonstrate PIWIMS performance in a 4-month real operating warehouse deployment and show that with only 32 manually labeled images per object, PIWIMS can achieve an accuracy of up to 87% in inventory tracking, which is a 28% increase when compared to traditional data augmentation techniques and 31% error reduction when compared to the third-party warehouse industry average. Furthermore, we demonstrate the ability of PI-WIMS to generalize across different camera angles and positions by achieving an accuracy of 85% in inventory tracking while varying the position and angle of the camera.

Real-world Machine Learning Systems: A survey from a Data-Oriented Architecture Perspective

“…Data that is produced by different sources in a continuous fashion fits the streams data model according to our review. That is the case of works that process video streams [43,49,56,60], sensors data [14,23,31,36,59,75,83,106,112,126], social media data [22,81,109,123,125], and network metrics [102,116]. Databases schemes are more appropriate as shared data models when system components require to store historical records [9,29,57] or process large amounts of data [105].…”

“…For example, systems can use API calls to collect data from data sources [10,81] or to receive requests from and send responses to end users [49] reviewed papers (i.e., 56.5%) do not adopt data coupling as data is hidden by each systems' components and these components interact using traditional process calls. They can use API calls [14,31,48,55,62,67,75,86,99,102,109] or remote procedure calls (RPC) [8,15,16,29,43,47,57,58,98,101,103,111,123,126] for communication. We found that not all the systems where components share a common data model necessarily follow the data coupling sub-principle.…”

“…Some of these systems are deployed on single cloud servers [10,31,47,49,56,81,98,105], where all data is stored. Other systems are deployed on resource-constraint devices like robots [57,67,106], or small processing units (e.g., Raspberry Pis) [23,43,62]. While cloud servers are flexible and can handle big-data requirements like the ones addressed by Habibi et al [56] or Schubert et al [105], resource-constraint devices limit the amount of stored data to the point where only trained learning models are deployed [62].…”

“…We found that just over half of the papers report systems where the whole decision-making process and the full learning models' life cycles happen in central nodes. Some of these works are implemented using traditional client-server architectures [10,16,47,49,56,81,98,105,116], others use sensors devices to perform basic data collection tasks [9,14,31,36,48,55,75,86,112], and others work in extreme resource constraint environments deploying the whole system in single and isolated devices [23,43,57,67,83,106].…”

“…The rest of the papers (i.e., 87%) do not follow this principle. Either they are implemented as centralised architectures [2,10,14,23,43,47,49,56,57,67,81,98,105,106,112,116], or federated architectures where there are no interactions between nodes at the same layer (e.g., servers at the edge layer) [8,9,16,29,31,36,48,55,58,62,66,75,83,86,99,[101][102][103].…”