Ahren Innovation Capital

Our co-founder, Prof. Shai Shalev-Shwartz, talks on Mentee’s Foundation Model, designed to let humans mentor robots through a combination of text instructions and real-time demonstrations—much like training a new employee on the job. The approach involves training a foundation model in simulation using reinforcement learning (RL) and applying Sim2Real techniques to transfer skills to physical robots. https://lnkd.in/dmUh6CNz #robotics #humanoid #humanoidrobotics #ai #reinforcementlearning

Today we’re announcing Heirloom's $150M Series B funding round! This investment, co-led by Future Positive and Lowercarbon Capital, brings together a coalition of global industrial leaders including Japan Airlines, Mitsubishi Corporation, Mitsui & Co., and Siemens Financial Services. Their participation reiterates our position as one of the leading developers of low-cost DAC technology. Why are industrial leaders backing Heirloom? Simple: we're showing a clear path to the lowest-cost DAC solution. Our approach combines abundant limestone (at $10-50/ton), smart automation (collecting 100M+ data points monthly), and modular design that enables manufacturing-style scale. Thank you to our investors, partners, and the incredible Heirloom team for making this possible. See some early coverage 👇 Financial Times ➡️ https://lnkd.in/dknfS2R3 Nikkei Asia ➡️ https://lnkd.in/ddDVJbwq Axios ➡️ https://lnkd.in/ddahNfwB Canary Media Inc. ➡️ https://lnkd.in/dZt5gzdt

MenteeBot is taking shopping assistance to the next level! In our latest end-to-end demo inspired by discussions with a major retailer, MenteeBot demonstrates its abilities as the ultimate shopping companion and highlights its practical application in real-world environments. As a leader in locomotion and dexterity, MenteeBot combines both in a robotic platform that can navigate a marker-less 3D world. Watch as MenteeBot grabs a shopping cart, follows a person in a wheelchair, and stops each time the person pauses to add items. The robot moves with a natural flow, making shopping more accessible and seamless. Our commitment to practical AI is evident in MenteeBot's seamless integration and natural interaction capabilities, refined inside the Isaac Lab simulator. Our technology bridges the gap between simulated environments and real-world applications, ensuring MenteeBot operates with unparalleled precision and adaptability. Stay tuned for MenteeBot V3, slated for release by late 2024, featuring detachable batteries to extend operational hours across diverse environments—from retail stores to manufacturing facilities and healthcare settings.

Really excited about our $40M+ Series A led by Khosla Ventures and Ahren Innovation Capital, and the support of leaders like Vinod Khosla Alexander A. Morgan, MD PhD, , Robert Gould, Alice Newcombe-Ellis, Joanna Green and others! These fundings support Rubedo transitioning to clinical stage our first program in the pipeline focused on chronic inflammatory skin conditions emerging from our discovery platform ALEMBIC with a first-in-class new molecule. Those skin conditions are large medical needs, typically aggravated in the elderly, and, regardless of the chronological age, they can manifest in any individual with a biologically aged skin. Skin is an entry point to then move beyond to other tissues and organs and other chronic diseases. Hevolution Foundation R42 Group Refactor Capital Knollwood Investment Advisory LongeVC Shanda Group Modi Ventures IAG

Today marks an exciting day as we finally launch Mentee Robotics to the public and unveil the MenteeBot, a humanoid robot that integrates AI across all its operational layers.   The prototype unveiled today, though not the final version, demonstrates for the first time a complete end-to-end cycle from a verbal command to complex task completion, including navigation, locomotion, scene understanding, object detection and localization and natural language understanding.   Locomotion, for example, is based on a novel Sim2Real machine-learning approach, wherein reinforcement learning occurs on a simulated version of the robot – thus utilizing unlimited data for training, followed by adaptation to the real world. The mapping of the environment is performed on the fly using NeRF-based algorithms. These cognitive maps store semantic information, allowing the robot to query them to find items and places. Transformer-based Large Language Models (LLMs) are used to interpret commands and “think through” the required steps to complete the task with an emphasis on the ability to integrate locomotion and dexterity (i.e., dynamically balancing the robot when carrying weights).   We are on the cusp of a convergence of computer vision, natural language understanding, strong and detailed simulators, and methodologies for transferring simulations to the real world. At Mentee Robotics, we see this convergence as the starting point for designing the future general-purpose robot that can move everywhere with the brains to perform household tasks and learn completely new tasks through imitation.   Mentee Robotics has operated in stealth mode for the past two years. Now, finally comes the time to showcase all that we have been working on. This is truly just the beginning

Ahren is proud to be partnering with Sunfire and an incredible group of co-investors in the company's €500M fundraise! This new round of financing will enable Sunfire to significantly increase the deployment of its electrolysis technologies, which are critical enablers for the decarbonization of hard-to-abate industries.