We present LOCUS (LOw-cost Customization for Universal Specialization), a pipeline that consumes few-shot data to streamline the construction and training of NLP models through targeted retrieval, synthetic data generation, and parameter-efficient tuning. With only a small number of labeled examples, LOCUS discovers pertinent data in a broad repository, synthesizes additional training samples via in-context data generation, and fine-tunes models using either full or low-rank (LoRA) parameter adaptation. Our approach targets named entity recognition (NER) and text classification (TC) benchmarks, consistently outperforming strong baselines (including GPT-4o) while substantially lowering costs and model sizes. Our resultant memory-optimized models retain 99% of fully fine-tuned accuracy while using barely 5% of the memory footprint, also beating GPT-4o on several benchmarks with less than 1% of its parameters.

In industrial scenarios, there is widespread use of collaborative robots (cobots), and growing interest is directed at evaluating and measuring the impact of some characteristics of the cobot on the human factor. In the present pilot study, the effect that the production rhythm (C1 - Slow, C2 - Fast, C3 - Adapted to the participant's pace) of a cobot has on the Experiential Locus of Control (ELoC) and the emotional state of 31 participants has been examined. The operators' performance, the degree of basic internal Locus of Control, and the attitude towards the robots were also considered. No difference was found regarding the emotional state and the ELoC in the three conditions, but considering the other psychological variables, a more complex situation emerges. Overall, results seem to indicate a need to consider the person's psychological characteristics to offer a differentiated and optimal interaction experience.

An important challenge for autonomous agents such as robots is to maintain a spatially and temporally consistent model of the world. It must be maintained through occlusions, previously-unseen views, and long time horizons (e.g., loop closure and re-identification). It is still an open question how to train such a versatile neural representation without supervision. We start from the idea that the training objective can be framed as a patch retrieval problem: given an image patch in one view of a scene, we would like to retrieve (with high precision and recall) all patches in other views that map to the same real-world location. One drawback is that this objective does not promote reusability of features: by being unique to a scene (achieving perfect precision/recall), a representation will not be useful in the context of other scenes. We find that it is possible to balance retrieval and reusability by constructing the retrieval set carefully, leaving out patches that map to far-away locations. Similarly, we can easily regulate the scale of the learned features (e.g., points, objects, or rooms) by adjusting the spatial tolerance for considering a retrieval to be positive. We optimize for (smooth) Average Precision (AP), in a single unified ranking-based objective. This objective also doubles as a criterion for choosing landmarks or keypoints, as patches with high AP. We show results creating sparse, multi-scale, semantic spatial maps composed of highly identifiable landmarks, with applications in landmark retrieval, localization, semantic segmentation and instance segmentation.

The last few years have been a major accelerator for the robotics industry at large, but warehouse robotics may be the biggest winner of all. Stay at home orders fueled adoption in the early days of the pandemic, as some retailers stayed open after being labeled "essential businesses." Even after things began reopening, those roles have remained difficult to fill, leading many firms to look toward robotic help. All the while, Amazon has had a jump on most of the industry, dating back to the company's acquisition of Kiva Systems a decade ago. The competition continues looking for angles to compete with the 800-pound e-commerce gorilla, and robotics startups have flooded the field, promising an edge.

Locus Robotics has signed what it claims to be one of the largest deployments of autonomous mobile robots (AMRs) ever. GEODIS is a leading global transport and logistics provider and has used Locus' AMRs since 2018, when it first deployed Locus' AMRs at a site in Indiana. The global third-party logistics company (3PL) has currently deployed Locus AMRs at 14 sites around the world, serving a variety of retail and consumer brands, including warehouses in the U.S and Europe. At press time, Locus hadn't provided how many of its AMRs GEODIS will have overall after these 1,000 are deployed. Locus told The Robot Report it doesn't have concrete evidence this is one of the largest AMR deals ever.

Place Recognition enables the estimation of a globally consistent map and trajectory by providing non-local constraints in Simultaneous Localisation and Mapping (SLAM). This paper presents Locus, a novel place recognition method using 3D LiDAR point clouds in large-scale environments. We propose a method for extracting and encoding topological and temporal information related to components in a scene and demonstrate how the inclusion of this auxiliary information in place description leads to more robust and discriminative scene representations. Second-order pooling along with a non-linear transform is used to aggregate these multi-level features to generate a fixed-length global descriptor, which is invariant to the permutation of input features. The proposed method outperforms state-of-the-art methods on the KITTI dataset. Furthermore, Locus is demonstrated to be robust across several challenging situations such as occlusions and viewpoint changes in 3D LiDAR point clouds. The open-source implementation is available at: https://github.com/csiro-robotics/locus .

Network-oriented research has been increasingly popular in many scientific areas. In neuroscience research, imaging-based network connectivity measures have become the key for understanding brain organizations, potentially serving as individual neural fingerprints. There are major challenges in analyzing connectivity matrices including the high dimensionality of brain networks, unknown latent sources underlying the observed connectivity, and the large number of brain connections leading to spurious findings. In this paper, we propose a novel blind source separation method with low-rank structure and uniform sparsity (LOCUS) as a fully data-driven decomposition method for network measures. Compared with the existing method that vectorizes connectivity matrices ignoring brain network topology, LOCUS achieves more efficient and accurate source separation for connectivity matrices using low-rank structure. We propose a novel angle-based uniform sparsity regularization that demonstrates better performance than the existing sparsity controls for low-rank tensor methods. We propose a highly efficient iterative Node-Rotation algorithm that exploits the block multi-convexity of the objective function to solve the non-convex optimization problem for learning LOCUS. We illustrate the advantage of LOCUS through extensive simulation studies. Application of LOCUS to Philadelphia Neurodevelopmental Cohort neuroimaging study reveals biologically insightful connectivity traits which are not found using the existing method.

June 2020 is off to a hot start for developers of autonomous mobile robots (AMRs). Yesterday, OTTO Motors announced a $29 million Series C, and today Locus Robotics closed $40 million in Series D funding. The Series D brings Locus' total amount of funding raised to $105 million. Locus' latest round was led by Zebra Ventures, the strategic investment arm of Zebra Technologies. Existing investors such as Scale Venture Partners also participated in the round.

How ready is India for the world of artificial intelligence (AI)? This question is answered in one of the latest global lists researched and created by Oxford Insights and commissioned by Canada's International Development Research Centre. It is called the Government Artificial Intelligence Readiness Index. The just released index measures 194 countries on a scale of 1-10 on how ready their governments are to embrace and make use of a world dominated by artificial intelligence. At the very top of the list is Singapore with a score of 9.186 and at the bottom is Somalia which scores 0.168. This kind of ranking is critical to understand the adoption of a technology which has, famously, been described as the'next electricity' or as fundamental as electricity by Andrew Ng, co-founder of Coursera and former head of Google Brain.

In the past decade, supply chain management has skyrocketed across verticals. With e-commerce becoming a critical economy booster, its success depends heavily on an efficient decision-making platform. With advanced technology solutions making steady inroads into businesses every day, the era of supply chain management has evolved. Several companies are thriving today as they capitalize on a robust technology architecture to support large-scale supply chain operations, and are providing analytical insights on route optimization freight tracking and analytics, sales beat optimization. Locus is a state-of-the-art decision-making platform for logistics, optimizing a range of operations to provide consistency, efficiency & transparency.