A new system combines simple control programs to enable fleets of robots — or other “multiagent systems” — to collaborate in unprecedented ways.
Writing a program to control a single autonomous robot navigating an uncertain environment with an erratic communication link is hard enough; write one for multiple robots that may or may not have to work in tandem, depending on the task, is even harder.
As a consequence, engineers designing control programs for “multiagent systems” — whether teams of robots or networks of devices with different functions — have generally restricted themselves to special cases, where reliable information about the environment can be assumed or a relatively simple collaborative task can be clearly specified in advance.
This May, at the International Conference on Autonomous Agents and Multiagent Systems, researchers from MIT’s Computer Science and Artificial Intelligence Laboratory (CSAIL) will present a new system that stitches existing control programs together to allow multiagent systems to collaborate in much more complex ways. The system factors in uncertainty — the odds, for instance, that a communication link will drop, or that a particular algorithm will inadvertently steer a robot into a dead end — and automatically plans around it.
For small collaborative tasks, the system can guarantee that its combination of programs is optimal — that it will yield the best possible results, given the uncertainty of the environment and the limitations of the programs themselves.
The Latest on: Multiagent systems
[google_news title=”” keyword=”Multiagent systems” num_posts=”10″ blurb_length=”0″ show_thumb=”left”]
via Google News
The Latest on: Multiagent systems
- 10 Free Courses to Build AI Agents in 2024on May 16, 2024 at 11:11 pm
A new era of autonomous AI agents has begun. The latest developments from Google I/O 2024 and OpenAI’s Spring Update confirm that all major companies are bullish on agents and that the future of AI ...
- Today’s AI models are impressive. Teams of them will be formidableon May 15, 2024 at 6:00 am
On May 13th OpenAI unveiled its latest model, GPT-4o. Mira Murati, the company’s chief technology officer, called it the “future of interaction between ourselves and the machines", because users can ...
- Best Multi-Line Business Phone Systemon May 14, 2024 at 5:00 pm
This guide was reviewed by a Business News Daily editor to ensure it provides comprehensive and accurate information to aid your buying decision. To find the best multiline phone systems for small ...
- Building Ethical AI: How To Take A User-Centered Approachon May 8, 2024 at 7:16 am
It’s crucial for tech companies to prioritize user well-being and ethical considerations when building or leveraging AI systems. Below, 20 members of Forbes Technology Council share strategies for ...
- AI Can Now Generate Entire Songs on Demand. What Does This Mean for Music as We Know It?on May 6, 2024 at 4:59 pm
The systems are very good at making up coherent songs that sound ... new interfaces for musical expression, and multi-agent models of social creativity. He is an associate professor at the School of ...
- An affordable miniature car-like robot to test control and estimation algorithmson April 29, 2024 at 10:56 am
which means that it can be adapted to test algorithms for various other systems, including model rockets and multi-agent robot teams. The researchers have already tested their hardware and ...
- Multiagent Systems Algorithmic, Game-Theoretic, and Logical Foundationson March 29, 2024 at 7:18 am
Multiagent systems combine multiple autonomous entities, each having diverging interests or different information. This overview of the field offers a computer science perspective, but also draws on ...
- Doctor of Philosophy in Software Engineeringon January 31, 2024 at 11:24 pm
Key research areas include requirements engineering, architecture, design, service-oriented computing, testing and verification, static analysis, software maintenance and multi-agent systems.
- Bo Yuon August 29, 2023 at 2:40 am
Dr. Yu’s research interests include decision making and support, multi-agent systems, market exchange mechanisms (negotiation and auction). He is also interested in design-type research and related ...
via Bing News
