"Rise of the Robots: The Competitive Pivot in AI Among Big Tech"
The race to dominate the Artificial Intelligence landscape is intensifying in the rapidly evolving world of technology. Big tech companies, including Microsoft, Amazon, and Alphabet, are not only competing against each other but also investing heavily in promising startups to accelerate their AI initiatives. The world's ten most valuable tech firms are all involved in developing technology for the robotics industry via acquisitions or partnerships. AI is fueling the rise of humanoid robots, creating an opportunity for tech giants. Humanoid robots are challenging to build and deploy because they require substantial sensor processing, advanced control, and complex skill execution. They also need multimodal artificial intelligence (AI) to understand the environment around them. Big tech companies see this challenge as an opportunity due to their software development advantages and capital deployment capabilities. For instance, Google and Nvidia have developed models and simulations to train robots, while Intel and Samsung have invested in humanoid robot startups. Additionally, Tesla has announced its humanoid robot, and other tech giants are also making their moves in this space. The technology industry is interconnected in many ways. For example, Microsoft's cloud services power OpenAI's agents, which are trained on Nvidia hardware fabricated in partnership with TSMC.
Let's look at some of the investments and organizational focus on robotics within tech companies.
Microsoft
The Autonomous Systems and Robotics Group at Microsoft researches and develops training and simulation technologies for robotics systems. Microsoft’s team collaborates closely with computer vision and reinforcement learning teams to conduct research at the intersection of large-scale multimodal representation learning, imitation learning, reinforcement learning, computer vision, and NLP. This enables them to train robots to perform new tasks naturally and effortlessly. One of the key challenges for robots is mastering human-level skills, particularly in manipulating objects. Therefore, the group's primary focus is on Project Dexter, aimed at addressing this challenge. It aims to make it possible to train robotic manipulation policies for real-world tasks in a cost-effective way in terms of expertise, time, and computational resources. In addition, Microsoft is making major investments in robotics. At the end of February this year, it spearheaded a massive $675 million Series B in Robotics startup Figure; other investors included Nvidia and Amazon.com founder Jeff Bezos. And last week, Microsoft announced a collaboration with Figure competitor Sanctuary AI, best known for its humanoid robot, Phoenix.
NVIDIA
NVIDIA has announced that it provides leading robot manufacturers, AI model developers, and software makers with various services, models, and computing platforms to accelerate global humanoid development. The offerings include new NVIDIA NIM™ microservices and frameworks for robot simulation and learning, the NVIDIA OSMO orchestration service for running multi-stage robotics workloads, and an AI- and simulation-enabled teleoperation workflow that enables developers to train robots using small amounts of human demonstration data.
NVIDIA offers a complete robotics platform for training, developing, and deploying AI-enabled robots on a large scale. Leading companies such as BYD Electronics, Siemens, Teradyne Robotics, and Intrinsic (an Alphabet company) are incorporating NVIDIA Isaac accelerated libraries, physically based simulation, and AI models into their software frameworks and robot models. This integration aims to improve efficiency and safety in factories, warehouses, and distribution centres and provide intelligent assistance for repetitive or highly precise tasks for human coworkers. The Isaac platform includes a variety of NVIDIA-accelerated libraries, AI foundation models, and simulation technologies that are currently accessible for robot makers to integrate into their technology stacks.
Tesla
In the news this week, Tesla announced that it is expanding its work on humanoid Optimus robots by hiring workers to collect data to train their AI. The company is looking for "data collection operators" who will wear motion-capture suits and virtual reality headsets to simulate the movements and actions of the robots. The job requires walking for over seven hours daily, and applicants must be between 5-foot-7 and 5-foot-11 to operate the motion-capture suits.
Tesla announced its Optimus project in 2021 to use it to complete factory tasks that were “unsafe, repetitive, or boring.” The initiative aligns with greater industrywide investment in automation spurred by pandemic-era labour shortages. As of 2016, 10% of warehouses reported using meaningful automation technology, according to Westernacher Consulting, with robot shipments expected to increase by up to 50% each year until 2030. The latest version of the robot completed its first independent task of managing batteries at a Tesla facility, as stated in its second-quarter earnings report for 2024. Elon Musk mentioned in July that the robots are expected to be in production for internal use as early as next year and to be available for sale to other companies by 2026. Tesla faces significant challenges if it aims to compete with other AI robotics companies. Competition in the field of task-automating robots is intensifying. With the successful funding round by Figure, the start-up specializing in AI-powered robotics, and OpenAI’s startup fund, Microsoft, Nvidia, and Jeff Bezos’s Bezos Expeditions, entering into a commercial agreement with BMW to support the carmaker in vehicle manufacturing. Figure and BMW will collaborate to identify the best applications for the technology and implement the robots at BMW's Spartanburg, S.C., factory.
Apple
According to Bloomberg's Mark Gurman, Apple focuses on robotics as its "next big thing." The company has engineering teams working on in-home robotic devices and the AI software and functionality that could make a home-helpful robot. Apple's robotics work is in the hardware engineering division, and its AI and machine learning group is under AI chief John Giannandrea.
Apple is moving forward with plans to create a tabletop robotic device as of August 2024. The company is prioritising the development of the device, overseen by Keven Lynch, who has also worked on the Apple Watch. Apple has already contacted suppliers for the tabletop robotic device.
The device has a sizeable iPad-like display mounted on a "thin robotic arm" that allows the display to tilt up and down and rotate 360 degrees. It is intended to serve as a "home command centre," a videoconferencing machine for FaceTime calls, and a home security monitoring tool. Apple plans to use Apple Intelligence and Siri for the robot, allowing it to respond to verbal commands, recognise different voices and orient the display to face users in the room. It is rumoured that Apple aims to launch the device in 2026 or 2027, with a target price of around $1,000.
Apple is also reportedly considering developing a mobile robot that can move around a user's home, potentially acting as a mobile iPad. This robot could make FaceTime calls, monitor the house and its occupants, perform simple tasks, and answer questions. Apple's interest in robotics comes from a desire to diversify its product offerings. Potential solutions to first-world problems could introduce a new personality akin to Siri. The shift towards hardware development reflects Apple's pursuit of new product categories for expansion. As a recent Bloomberg newsletter explained, Apple already plays a significant role in most customers' lives with its products, such as iPhones, Macs, wearables, and Apple TV. While the company works to enhance its existing hardware, the challenge lies in introducing a new product in a similar field that is innovative and distinct enough to be worthwhile.
Amazon
Amazon is selling a $1,600 "Astro" robot by invitation. Astro can navigate the home and check specific rooms, people, and things remotely when the user is away. The robot can send an alert if it detects an unrecognized person or certain sounds like an alarm. Astro has a built-in display and can be used for watching TV, making calls, setting reminders, sending messages, and more. It uses Alexa and Amazon suggests it can be used to "remotely care for aging loved ones." Amazon's Innovation Lab, which is home to a diverse team of scientists and engineers from around the world, is developing and testing new technologies to support employees better and serve customers.
Nestled among the centuries-old rice fields of Vercelli, northern Italy, is a building that is shaping the future of work. The Amazon Operations Innovation Lab is one of Europe’s most advanced centres for robotic innovation. It serves as an international hub for scientists and engineers to develop and test new technologies, such as advanced robotics and AI-powered innovations. The lab, established in 2017, is the central base for the global Mechatronics & Sustainable Packaging team, which was formed in 2019. By the end of 2024, this team will have installed over 1,000 new robotics and AI-powered innovations throughout Amazon’s European fulfilment centre (FC) network, representing an investment of over €700 million. These cutting-edge technologies include item sorters, pallet movers, and automated guide vehicles, which actively support Amazon employees in their roles across various sites. Robotics and technology are paramount in enhancing Amazon's operations network, while AI enables the scaling of these advancements for a global operation serving millions of customers worldwide.
With Big Tech companies increasing investments in robotics, the future of technology and industry is poised for transformative changes. The integration of AI in robotic systems will lead to more autonomous robots, enhancing productivity and safety. Robotics is expected to significantly impact daily life, and ethical considerations will become increasingly important. Advancements in materials science will lead to more sophisticated, lightweight, and durable robots. The next phase of robotic development will focus on greater AI integration, collaborative functionalities, everyday utility, and responsible innovation.
