Researchers are developing better, faster ways of providing human guidance to autonomous robots.
Told to optimize for speed while racing down a track in a computer game, a car pushes the pedal to the metal … and proceeds to spin in a tight little circle. Nothing in the instructions told the car to drive straight, and so it improvised.
This example – funny in a computer game but not so much in life – is among those that motivated Stanford University researchers to build a better way to set goals for autonomous systems.
Dorsa Sadigh, assistant professor of computer science and of electrical engineering, and her lab have combined two different ways of setting goals for robots into a single process, which performed better than either of its parts alone in both simulations and real-world experiments. The researchers presented the work June 24 at the Robotics: Science and Systems conference.
“In the future, I fully expect there to be more autonomous systems in the world and they are going to need some concept of what is good and what is bad,” said Andy Palan, graduate student in computer science and co-lead author of the paper. “It’s crucial, if we want to deploy these autonomous systems in the future, that we get that right.”
The team’s new system for providing instruction to robots – known as reward functions – combines demonstrations, in which humans show the robot what to do, and user preference surveys, in which people answer questions about how they want the robot to behave.
“Demonstrations are informative but they can be noisy. On the other hand, preferences provide, at most, one bit of information, but are way more accurate,” said Sadigh. “Our goal is to get the best of both worlds, and combine data coming from both of these sources more intelligently to better learn about humans’ preferred reward function.”
Demonstrations and surveys
In previous work, Sadigh had focused on preference surveys alone. These ask people to compare scenarios, such as two trajectories for an autonomous car. This method is efficient, but could take as much as three minutes to generate the next question, which is still slow for creating instructions for complex systems like a car.
To speed that up, the group later developed a way of producing multiple questions at once, which could be answered in quick succession by one person or distributed among several people. This update sped the process 15 to 50 times compared to producing questions one-by-one.
The new combination system begins with a person demonstrating a behavior to the robot. That can give autonomous robots a lot of information, but the robot often struggles to determine what parts of the demonstration are important. People also don’t always want a robot to behave just like the human that trained it.
“We can’t always give demonstrations, and even when we can, we often can’t rely on the information people give,” said Erdem Biyik, a graduate student in electrical engineering who led the work developing the multiple-question surveys. “For example, previous studies have shown people want autonomous cars to drive less aggressively than they do themselves.”
That’s where the surveys come in, giving the robot a way of asking, for example, whether the user prefers it move its arm low to the ground or up toward the ceiling. For this study, the group used the slower single question method, but they plan to integrate multiple-question surveys in later work.
In tests, the team found that combining demonstrations and surveys was faster than just specifying preferences and, when compared with demonstrations alone, about 80 percent of people preferred how the robot behaved when trained with the combined system.
“This is a step in better understanding what people want or expect from a robot,” said Sadigh. “Our work is making it easier and more efficient for humans to interact and teach robots, and I am excited about taking this work further, particularly in studying how robots and humans might learn from each other.”
Better, faster, smarter
People who used the combined method reported difficulty understanding what the system was getting at with some of its questions, which sometimes asked them to select between two scenarios that seemed the same or seemed irrelevant to the task – a common problem in preference-based learning. The researchers are hoping to address this shortcoming with easier surveys that also work more quickly.
“Looking to the future, it’s not 100 percent obvious to me what the right way to make reward functions is, but realistically you’re going to have some sort of combination that can address complex situations with human input,” said Palan. “Being able to design reward functions for autonomous systems is a big, important problem that hasn’t received quite the attention in academia as it deserves.”
The team is also interested in a variation on their system, which would allow people to simultaneously create reward functions for different scenarios. For example, a person may want their car to drive more conservatively in slow traffic and more aggressively when traffic is light.
The Latest on: Autonomous systems
via Google News
The Latest on: Autonomous systems
- Smooth touchdown: novel camera-based system for automated landing of drone on a fixed spoton January 23, 2021 at 1:24 pm
"While it is desirable to automate the landing using a depth camera that can gauge terrain unevenness and find suitable landing spots, a framework serving as a useful base needs to be developed first, ...
- Waymo CEO Says Tesla Is Not A Competitor, Gives Estimated Cost Of Autonomous Vehicleson January 22, 2021 at 7:01 pm
In an interview with German business publication Manager Magazin, Waymo CEO John Krafcik gave an idea of how much each one of the company’s vehicles will cost.
- Autonomous Ships Market by Autonomy, Ship Type, End-use, Solution, Propulsion and Region - Global Forecast to 2030 - ResearchAndMarkets.comon January 22, 2021 at 9:25 am
The "Autonomous Ships Market by Autonomy (Fully Autonomous, Remote Operations, Partial Automation), Ship Type (Commercial, Defense), End-Use (Linefit, Retrofit), Solution (Systems, Software, ...
- XPeng to deploy Amap’s 3rd-gen in-car navigation systemon January 22, 2021 at 5:24 am
XPeng Inc., a leading Chinese smart electric vehicle company, today announced that it will be the world’ s first auto maker to deploy Amap’ s 3rd- generation in-car navigation system. The XPeng P7 ...
- Autonomous System for Ships Tested at Antwerp Porton January 22, 2021 at 12:32 am
Recently, researchers from the University of Antwerp have developed a 3D sonar sensor for ships' autonomous sailing. How does the new sensor work, has the new sensor been tested, and can it help the ...
- Global Autonomous Ships Market (2020 to 2030) - Increasing Use of Automated Systems to Reduce Human Errors and Risks is Driving Growthon January 21, 2021 at 7:56 pm
The autonomous ships market is estimated to be USD 5,866 million in 2020 and is projected to reach USD 14,256 million by 2030, at a CAGR of 9.3% from 2020 to 2030. The demand for new ships and the ...
- New Robotics & Autonomous Systems Master’s Draws Strong Intereston January 20, 2021 at 4:00 pm
New Robotics & Autonomous Systems Master’s Draws Strong Interest. in NEWS, Spotlight Master's. Students, companies eager to participate. By Michael Seele. Start a new master’s ...
- Use of autonomous vehicles in mining and farming touted at CES 2021on January 20, 2021 at 10:23 am
Caterpillar and John Deere showcased autonomous machines that are being used worldwide in farming and mining projects.
- Toyota builds an autonomous drift Supra to develop advanced driver assistance systemson January 19, 2021 at 9:47 am
Toyota has revealed an autonomous Supra that drifts. It's meant to help researchers develop driver assistance tech.
- Dragontail Systems Deploys Autonomous Drones To Assist Food Deliverieson January 19, 2021 at 6:00 am
Dragontail Systems Limited (ASX: DTS), a software company that optimizes the entire restaurant order and delivery process for global ...
via Bing News