A version of this questionnaire first appeared in TechCrunch’s free robotics newsletter, Actuator. Register here.
We wrap up our year-end robotics Q&A series with this entry from Deepu Talla. We when I visited NVIDIA’s Bay Area headquarters in October. For more than a decade, Talla was the chip giant’s Vice President and General Manager – Embedded & Edge Computing. It offers a unique insight into the state of robotics in 2023 and where things are headed in the future. In recent years, NVIDIA has established a major platform for robotics simulation, prototyping, and development.
Previous Questions & Answers:
Image Credits: NVIDIA
What role(s) will genetic artificial intelligence play in the future of robotics?
We are already seeing productivity improvements with genetic AI across industries. Clearly, the impact of GenAI will be transformational in robotics from simulation to design and more.
- Simulation: Modelers will be able to accelerate simulation development, bridging the gaps between 3D technical artists and developers, creating scenes, building environments, and creating assets. These elements of GenAI will see increased use for synthetic data generation, robot skill training, and software testing.
- Multimodal AI: Transformer-based models will improve robots’ ability to better understand the world around them, allowing them to work in more environments and complete complex tasks.
- Robot (re)programming: Greater ability to define tasks and functions in plain language to make robots more general/multipurpose.
- Design: New mechanical designs for better performance — for example, end-effectors.
What are your thoughts on the humanoid form factor?
Designing autonomous robots is difficult. Humanoids are even tougher. Unlike most AMRs that primarily understand floor-level obstacles, humanoids are mobile operators that will need multimodal AI to understand more of the environment around them. An incredible amount of sensor processing, advanced control and execution skills are required.
Innovations in AI production capabilities to create fundamental models make the robot skills required for humanoids more generalizable. At the same time, we are seeing advances in simulations that can train AI-based control systems as well as perception systems.
After manufacturing and warehouses, what is the next big category for robotics?
Markets where businesses are feeling the effects of labor shortages and demographic changes will continue to align with corresponding robotics opportunities. This covers robotics companies working in a variety of industries, from agriculture to last-mile delivery to retail and more.
A key challenge in building autonomous robots for different classes is building the 3D virtual worlds needed to simulate and test the stacks. Again, genetic AI will help by allowing developers to more quickly create realistic simulation environments. The integration of artificial intelligence into robotics will enable increased automation in more active and less robot-friendly environments.
How far are true general purpose robots?
We continue to see robots become more intelligent and capable of multitasking in a given environment. We expect to see continued focus on mission-related problems while making them more generalizable. True general-purpose built-in autonomy is further out.
Will home robots take off (beyond the gaps) in the next decade?
We will have helpful personal assistants, lawnmowers and robots to help seniors share.
The trade-off holding back home robots, to date, is the pivot of how much someone is willing to pay for their robot and whether the robot delivers that value. Robot vacuums have long delivered value for their price, hence their popularity.
Also, as bots become smarter, having intelligent user interfaces will be key to increased adoption. Robots that can map their own environment and receive instructions via speech will be easier for home consumers to use than robots that require some programming.
The next category to take off will likely focus first on outdoor spaces—for example, autonomous lawn care. Other home robots such as personal/healthcare assistants show great promise, but must address some of the indoor challenges encountered in dynamic, unstructured home environments.
What major robotics story/trend isn’t getting enough coverage?
The need for a platform approach. Many robotics startups fail to scale as they build robots that work well for a specific task or environment. For commercial viability at scale, it is important to develop robots that are more generalizable — that is, they can add new skills quickly or bring existing skills to new environments.
Roboticists need platforms with tools and libraries to train and test AI for robotics. The platform should provide simulation capabilities to train models, generate synthetic data, and exercise the entire robotics software stack, with the ability to run the latest and emerging AI models directly on the robot.
Tomorrow’s successful robotics startups and companies will need to focus on developing new robot skills and automating tasks and leverage the full extent of available end-to-end development platforms.
