Urban transportation systems encounter diverse challenges across multiple tasks, such as traffic forecasting, electric vehicle (EV) charging demand prediction, and taxi dispatch. Existing approaches suffer from two key limitations: small-scale deep learning models are task-specific and data-hungry, limiting their generalizability across diverse scenarios, while large language models (LLMs), despite offering flexibility through natural language interfaces, struggle with structured spatiotemporal data and numerical reasoning in transportation domains. To address these limitations, we propose TransLLM, a unified foundation framework that integrates spatiotemporal modeling with large language models through learnable prompt composition. Our approach features a lightweight spatiotemporal encoder that captures complex dependencies via dilated temporal convolutions and dual-adjacency graph attention networks, seamlessly interfacing with LLMs through structured embeddings. A novel instance-level prompt routing mechanism, trained via reinforcement learning, dynamically personalizes prompts based on input characteristics, moving beyond fixed task-specific templates. The framework operates by encoding spatiotemporal patterns into contextual representations, dynamically composing personalized prompts to guide LLM reasoning, and projecting the resulting representations through specialized output layers to generate task-specific predictions. Experiments across seven datasets and three tasks demonstrate the exceptional effectiveness of TransLLM in both supervised and zero-shot settings. Compared to ten baseline models, it delivers competitive performance on both regression and planning problems, showing strong generalization and cross-task adaptability. Our code is available at https://github.com/BiYunying/TransLLM.

Urban transportation plays a vital role in modern city life, affecting how efficiently people and goods move around. This study analyzes transportation patterns using two datasets: the NYC Taxi Trip dataset from New York City and the Pathao Food Trip dataset from Dhaka, Bangladesh. Our goal is to identify key trends in demand, peak times, and important geographical hotspots. We start with Exploratory Data Analysis (EDA) to understand the basic characteristics of the datasets. Next, we perform geospatial analysis to map out high-demand and low-demand regions. We use the SARIMAX model for time series analysis to forecast demand patterns, capturing seasonal and weekly variations. Lastly, we apply clustering techniques to identify significant areas of high and low demand. Our findings provide valuable insights for optimizing fleet management and resource allocation in both passenger transport and food delivery services. These insights can help improve service efficiency, better meet customer needs, and enhance urban transportation systems in diverse urban environments.

Unmanned aerial vehicles (UAV) showed great potential in improving the efficiency of parcel delivery applications in the coming smart cities era. Unfortunately, the trustworthy positioning and control algorithms of the UAV are significantly challenged in complex urban areas. For example, the ubiquitous global navigation satellite system (GNSS) positioning can be degraded by the signal reflections from surrounding high-rising buildings, leading to significantly increased positioning uncertainty. An additional challenge is introduced to the control algorithm due to the complex wind disturbances in urban canyons. Given the fact that the system positioning and control are highly correlated with each other, for example, the system dynamics of the control can largely help with the positioning, this paper proposed a joint positioning and control method (JPCM) based on factor graph optimization (FGO), which combines sensors' measurements and control intention. In particular, the positioning measurements are formulated as the factors in the factor graph model, such as the positioning from the GNSS. The model predictive control (MPC) is also formulated as the additional factors in the factor graph model. By solving the factor graph contributed by both the positioning factor and the MPC-based factors, the complementariness of positioning and control can be fully explored. To guarantee reliable system dynamic parameters, we validate the effectiveness of the proposed method using a simulated quadrotor system which showed significantly improved trajectory following performance. To benefit the research community, we open-source our code and make it available at https://github.com/RoboticsPolyu/IPN_MPC.

As head strategic advisor, AI Venturetech has been instrumental in guiding TulsaLabs, a division of Appswarm (OTC:SWRM), in building their next generation technology research lab in Tulsa, OK. Although still in development stage, TulsaLabs has moved forward with its first clients to begin expanding its technological reach into areas such as wearables, AI, and new urban transportation. TulsaLabs will utilize data generated from GTXO wearable devices and sensors to develop cloud-based analytics that will monitor and analyze a user's health data through long-term dataset development utilizing artificial intelligence. This platform data will allow users to detect potential trends that can be offset with lifestyle adjustments to prevent the onset of some of the most common metabolic diseases such as diabetes, heart disease, and obesity. Combined together, GTXO and SWRM look to break into the exciting life extension market as emerging player in the wearable industry.

NEW YORK, September 23, 2021--(BUSINESS WIRE)--Helbiz Inc. (NASDAQ: HLBZ), a global leader in micro-mobility and the first in its industry to be publicly listed on Nasdaq, today announced a partnership with Drover AI to integrate its PathPilot safety technology onto Helbiz e-scooters. Helbiz will be the exclusive operator of PathPilot in Italy, with an initial deployment in Milan by end of the year. The company plans to expand the integration across other markets as the partnership grows. This press release features multimedia. The PathPilot technology is powered by artificial intelligence and computer vision, using onboard cameras to locate the surroundings of e-scooters.

Welcome to California Inc., the weekly newsletter of the L.A. Times Business Section. One story stood out for me: Uber says it will introduce flying taxis in Los Angeles by 2020. "We're trying to work with cities in the early days who are interested in partnering to make it happen, while knowing that there will be pitfalls along the way," says Jeff Holden, Uber's chief product officer. Solar panels: President Trump will be presented with a plan Monday to impose restrictions and tariffs on imports of the most popular photovoltaic generating panels used in the booming U.S. residential and utility-scale solar markets. The U.S. International Trade Commission, an independent agency, has proposed the action after two international solar-panel producers with U.S. plants complained that they needed protection from low-cost imports.

Sparks of electricity emanating from a Tesla coil at the Mendeleyevskaya metro station in Moscow, Russia, January 24, 2016. Editor's Note: This is an excerpt from John Wasik's new book, "Lightning Strikes: Timeless Lessons in Creativity from the Life and Work of Nikola Tesla" (Sterling, 2016), slightly edited for this column. World-changing inventions made Nikola Tesla a celebrity in his own time, but something otherworldly makes him transcend his era and remain a perpetual beacon for our civilization 70 years after his death. He's now an immortal rock star, an icon for billionaires, cyberpunks, artists and "maker" inventors who are still fiddling with everyday machines in their basements and garages. Search engine designers, energy czars, musicians, artists and creators everywhere feel his influence.