When minimizing makespan during off-line planning, the fastest action sequence to reach a particular state is, by definition, preferred. When trying to reach a goal quickly in on-line planning, previous work has inherited that assumption: the faster of two paths that both reach the same state is usually considered to dominate the slower one. In this short paper, we point out that, when planning happens concurrently with execution, selecting a slower action can allow additional time for planning, leading to better plans. We present Slo'RTS, a metareasoning planning algorithm that estimates whether the expected improvement in future decision-making from this increased planning time is enough to make up for the increased duration of the selected action. Using simple benchmarks, we show that Slo'RTS can yield shorter time-to-goal than a conventional planner. This generalizes previous work on metareasoning in on-line planning and highlights the inherent uncertainty present in an on-line setting.

The 119 drones underwent catapult-assisted take-offs and performed aerial formations, the Xinhua News Agency reported on Sunday. The CETC said "swarm intelligence" is regarded as the core of the artificial intelligence of unmanned systems and the future of intelligent unmanned systems. The huge scale of low cost and multi-function UAVs could be used in risky tasks such as emergency communications. CETC engineer Zhao Yanjie said since drones were invented in 1917, intelligent swarms have "changed the rules of the game." In November 2016, the CETC launched 67 drones during the China International Aviation & Aerospace Exhibition in Zhuhai, South China's Guangdong Province, breaking the previous record of 50 drones by the US Navy, CCTV reported.

In multi-agent path finding (MAPF) the task is to find non-conflicting paths for multiple agents. Recently, a SAT-based approach was developed to solve this problem and proved beneficial in many cases when compared to other search-based solvers. In this paper, we introduce SAT-based unbounded- and bounded-suboptimal algorithms and compare them to relevant search-based algorithms.

In the multi-agent path-finding (MAPF) problem a plan is needed to move a set of agents from their initial location to their goals without collisions. In this paper we introduce and study the k -robust MAPF problem, where we seek a plan that is robust to k unexpected delays per agent. We say that a plan π is k -robust if it does not have any k - delay conflicts. Informally, this means that no conflicts will occur even if some of the agents are delayed by up to k time steps. The problem we address in this paper is how to find optimal sum-of-costs k -robust plans.

Multi-agent pathfinding (MAPF) is an area of expanding research interest. At the core of this research area, numerous diverse search-based techniques were developed in the past 6 years for optimally solving MAPF under the sum-of-costs objective function. In this paper we survey these techniques, while placing them into the wider context of the MAPF field of research. Finally, we provide analytical and experimental comparisons that show that no algorithm dominates all others in all circumstances. We conclude by listing important future research directions.

A hybrid observer is described for estimating the state of an $m>0$ channel, $n$-dimensional, continuous-time, distributed linear system of the form $\dot{x} = Ax,\;y_i = C_ix,\;i\in\{1,2,\ldots, m\}$. The system's state $x$ is simultaneously estimated by $m$ agents assuming each agent $i$ senses $y_i$ and receives appropriately defined data from each of its current neighbors. Neighbor relations are characterized by a time-varying directed graph $\mathbb{N}(t)$ whose vertices correspond to agents and whose arcs depict neighbor relations. Agent $i$ updates its estimate $x_i$ of $x$ at "event times" $t_1,t_2,\ldots $ using a local observer and a local parameter estimator. The local observer is a continuous time linear system whose input is $y_i$ and whose output $w_i$ is an asymptotically correct estimate of $L_ix$ where $L_i$ a matrix with kernel equaling the unobservable space of $(C_i,A)$. The local parameter estimator is a recursive algorithm designed to estimate, prior to each event time $t_j$, a constant parameter $p_j$ which satisfies the linear equations $w_k(t_j-\tau) = L_kp_j+\mu_k(t_j-\tau),\;k\in\{1,2,\ldots,m\}$, where $\tau$ is a small positive constant and $\mu_k$ is the state estimation error of local observer $k$. Agent $i$ accomplishes this by iterating its parameter estimator state $z_i$, $q$ times within the interval $[t_j-\tau, t_j)$, and by making use of the state of each of its neighbors' parameter estimators at each iteration. The updated value of $x_i$ at event time $t_j$ is then $x_i(t_j) = e^{A\tau}z_i(q)$. Subject to the assumptions that (i) the neighbor graph $\mathbb{N}(t)$ is strongly connected for all time, (ii) the system whose state is to be estimated is jointly observable, (iii) $q$ is sufficiently large, it is shown that each estimate $x_i$ converges to $x$ exponentially fast as $t\rightarrow \infty$ at a rate which can be controlled.

Many real-world problems, such as network packet routing and urban traffic control, are naturally modeled as multi-agent reinforcement learning (RL) problems. However, existing multi-agent RL methods typically scale poorly in the problem size. Therefore, a key challenge is to translate the success of deep learning on single-agent RL to the multi-agent setting. A major stumbling block is that independent Q-learning, the most popular multi-agent RL method, introduces nonstationarity that makes it incompatible with the experience replay memory on which deep Q-learning relies. This paper proposes two methods that address this problem: 1) using a multi-agent variant of importance sampling to naturally decay obsolete data and 2) conditioning each agent's value function on a fingerprint that disambiguates the age of the data sampled from the replay memory. Results on a challenging decentralised variant of StarCraft unit micromanagement confirm that these methods enable the successful combination of experience replay with multi-agent RL.

Multiagent environments where agents compete for resources are stepping stones on the path to AGI. Multiagent environments have two useful properties: first, there is a natural curriculum -- the difficulty of the environment is determined by the skill of your competitors (and if you're competing against clones of yourself, the environment exactly matches your skill level). Second, a multiagent environment has no stable equilibrium: no matter how smart an agent is, there's always pressure to get smarter. These environments have a very different feel from traditional environments, and it'll take a lot more research before we become good at them. We've developed a new algorithm, MADDPG, for centralized learning and decentralized execution in multiagent environments, allowing agents to learn to collaborate and compete with each other.

Freelance workers enable the automation of their work -- ironically. AGI (artificial general intelligence) is the pursuit of machine cognition, largely still a work in progress. Pat Langley from Arizona State University, has an excellent essay highlighting the differences between most of what's labeled "AI", and what he refers to as the'Cognitive Systems Paradigm'. A cognitive system has the machinery to begin working with written language, use heuristics and other approaches to deal with incomplete data, make inferences from structured representations of information, and so on. Some current AGI frameworks can productively be part of this type of work, although unfortunately many projects are not yet well documented or openly available.

In a winning scenario, smiling can decrease your odds of success against the same opponent in subsequent matches, according to new research presented by the USC Institute for Creative Technologies and sponsored by the U.S. Army Research Laboratory. People who smiled during victory increased the odds of their opponent acting aggressively to steal a pot of money rather than share it in future gameplay, according to a paper presented in May at the International Conference on Autonomous Agents and Multiagent Systems by USC ICT research assistant Rens Hoegen, USC ICT research programmer Giota Stratou and Jonathan Gratch, director of virtual humans research at USC ICT and a professor of computer science at the USC Viterbi School of Engineering. Conversely, researchers found smiling during a loss tended to help the odds of success in the game going forward. The study is in line with previous research published by senior author Gratch, whose main interest lies both in how people express these tells -- an unconscious action that betrays deception -- and using this data to create artificial intelligence to discern and even express these same emotional cues as a person. "We think that emotion is the enemy of reason. But the truth is that emotion is our way of assigning value to things," said Gratch.