Computing Reviews
Today's Issue Hot Topics Search Browse Recommended My Account Log In
Review Help
Search
Deliberation for autonomous robots
Ingrand F., Ghallab M. Artificial Intelligence247  10-44,2017.Type:Article
Date Reviewed: Jul 14 2017

Autonomous robots use sensors and actuators to gain information about their environments and act on the information to achieve assigned goals and missions. This paper presents a survey on the topic of deliberation in autonomous robots.

Deliberation functions enable robots to behave intelligently, like they are thinking by predicting effects of their actions and making proper decisions about their courses of actions. The deliberation functions mentioned in the survey are planning, acting, monitoring, observing, goal reasoning, and learning.

Deliberation is critical when it is unfeasible or impractical to conceive of all possible courses of action in advance and program the robots to perform autonomously. This is the case in situations where autonomous robots are to operate in diverse environments, perform a variety of tasks, and deal with a range of interactions. Deliberation functions are embedded in many unmanned systems including unmanned aerial vehicles (UAVs); autonomous underwater vehicles (AUVs); self-driving cars; and many industrial, service, or field applications.

The survey is motivated by what the authors say is a lack of “global” perspective on deliberation functions in robotics, on the challenges they pose, and on the techniques needed for their design and implementation.

The authors make three disclaimers: they do not propose a unifying theory of deliberation; they do not survey human-robot and multirobot interaction and cooperation issues; and they focus mainly on planning and acting and much less on observing, monitoring, goal reasoning, and learning.

Section 1 (“Introduction”) gives general motivation for the survey and mentions that deliberation is at the intersection of robotics and artificial intelligence.

Section 2 (“Design of a robotics deliberation system”) characterizes robotic applications along a number of loosely defined features, reviews design options of a deliberation system, and gives a global view of surveyed systems summarized in two tables (Tables 2 and 3).

Section 3 (“Planning”) reviews options for the design of a robot planner and summarizes in Table 4 the features of the approaches used in surveyed systems. Section 3 describes various approaches suited for classes of planning systems: motion and task planning, temporal planning, probabilistic planning, and planning in open domains.

Section 4 (“Acting”) describes the acting carried-out functions and summarizes in Table 6 the approaches used in the surveyed systems to design a deliberate acting system.

Section 5 (“Monitoring and goal reasoning”) summarizes in Table 7 the approaches used in surveyed systems and discusses the approaches of model-based diagnosis, plan-invariance, temporal action logic, and goal reasoning.

Section 6 (“Observing”) summarizes in Table 8 the approaches and main functions of surveyed observing systems: planning to perceive, object recognition and symbol anchoring, event and situation recognition, and stream-based integration.

Section 7 (“Acquiring deliberation models”) reviews two approaches to the acquisition of models needed for and specific to deliberation: acquiring semantic maps and acquiring planning and acting models.

Section 8 contains general discussions and an attempt to synthesize an overview of the state of the art on deliberation in robotics.

Section 9 gives concluding remarks, a take-home message, and recommendations for future research directions to address representation, modeling, and integration issues.

The paper is a bit tedious to read, as it appears to be a collection of disparate contents with few unifying themes or binding threads. Also the lack of any formalization makes it difficult not to get lost in lengthy narrations. The take-home message stated in the introduction and repeated in the conclusion is not explicitly advocated by the content of the paper. The authors also do not provide strong arguments or empirical evidence in support of postulates and claims. An example is the claim that “deliberation can endow a robotic system with extended, more flexible and robust functionalities; it can reduce its deployment cost.” Overall, the proposed global view of surveyed systems does little to contribute to the aim of helping with “perceiving the forest from the trees” in this area of active research.

Reviewer:  Yousri El Fattah Review #: CR145425 (1709-0634)
Bookmark and Share
  Featured Reviewer  
 
Robotics (I.2.9 )
 
 
Workcell Organization And Planning (I.2.9 ... )
 
 
Learning (I.2.6 )
 
Would you recommend this review?
yes
no
Other reviews under "Robotics": Date
Movement problems for 2-dimensional linkages
Hopcroft J. (ed), Joseph D., Whitesides S. SIAM Journal on Computing 13(3): 610-629, 1984. Type: Article
Feb 1 1985
Robot motion planning with uncertainty in control and sensing
Latombe J. (ed), Lazanas A., Shekhar S. Artificial Intelligence 52(1): 1-47, 1991. Type: Article
Oct 1 1992
Dictionary of robot technology in four languages: English, German, French, Russian
Bürger E., Korzak G., Elsevier North-Holland, Inc., New York, NY, 1986. Type: Book (9789780444995193)
Mar 1 1988
more...

E-Mail This Printer-Friendly
Send Your Comments
Contact Us
Reproduction in whole or in part without permission is prohibited.   Copyright 1999-2024 ThinkLoud®
Terms of Use
| Privacy Policy