Rule Book - robocupatwork.orgrobocupatwork.org/download/rulebook-2015-06-04.pdf · Rule Book Version 2015.2 Juni 4, 2015 Status: China 2015 Version Authors: Walter Nowak Gerhard Kraetzschmar

1

Rule Book

Version 2015.2Juni 4, 2015

Status: China 2015 Version

Authors:

Walter NowakGerhard KraetzschmarNico Hochgeschwender

Rainer BischoffDaniel Kaczor

Frederik HeggerJan Carstensen

RoboCup@Work Rule Book

2

Table of Contents1. RoboCup @ Work in a Nutshell2. Organization of the League

2.1 League Committees2.1.1 Technical Committee2.1.2 Organizing Committee2.1.3 Industrial Advisory Board2.1.4 Research Advisory Board

2. 2 League Infrastructure2.2.1 Infrastructure Maintained by the League2.2.2 Infrastructure Maintained by Teams

3. Participation in the Competition3.1 Steps to Participate3.2 Qualification3.3 Team Description Paper

4. Organization of the Competition4.1 Teams4.2 Team Practice and Use of Arenas4.3 Stages and Tests4.4 Scoring and Ranking4.5 Common Procedures4.6 Roles of OC and TC4.7 Referees4.8 Meetings and Language of Communication4.9 Code of Conduct and Disqualification

5. General Design of RoboCup @ Work Scenarios5.1 Assumptions about Robots used in the Competition5.2 Rules Applying to Robots

5.2.1 Robot Design and Design Constraints5.2.2 Robot Behavior and Safety

5.3 Use of External Devices5.4 Design of Competition Arenas

5.4.1 Size of the Arena5.4.2 Floor5.4.3 Walls5.4.4 Service Areas5.4.5 Racks5.4.6 Shelves5.4.7 Places5.4.8 Obstacles5.4.9 Floor Markers

5.5 Design of Manipulation Tasks5.5.1 Manipulation Objects5.5.2 Placement of objects5.5.3 Grasping objects5.5.4 Dropping objects

5.6 Overall Scenario and Variation Points6. Tests


3

6.1 Basic Navigation Test ( BNT )6.1.1 Purpose and Focus of the Test6.1.2 Scenario Environment6.1.3 Manipulation Objects6.1.4 Task6.1.5 Complexity Levels6.1.6 Rules6.1.7 Scoring

6.2 Basic Manipulation Test ( BMT )6.2.1 Purpose and Focus of the Test6.2.2 Scenario Environment6.2.3 Manipulation Objects6.2.4 Task6.2.6 Rules6.2.7 Scoring

6.3 Basic Transportation Test ( BTT )6.3.1 Purpose and Focus of the Test6.3.2 Scenario Environment6.3.3 Manipulation Objects6.3.4 Task6.3.6 Rules6.3.7 Scoring

6.4 Conveyor Belt Test ( CBT )6.4.1 Purpose and Focus of the Test6.4.2 Scenario Environment6.4.3 Manipulation Objects6.4.4 Task6.4.6 Rules6.4.7 Scoring

6.5 Precision Placement Test ( PPT )6.5.1 Purpose and Focus of the Test6.5.2 Scenario Environment6.5.3 Manipulation Objects6.5.4 Task6.5.5 Complexity Levels6.5.6 Rules6.5.7 Scoring

6.5 Final ( Example )6.6.1 Purpose and Focus of the Test6.6.2 Scenario Environment6.6.3 Complexity Levels6.6.4 Scoring


4

1. RoboCup@Work in a NutshellRoboCup@Work is a new competition in RoboCup that targets the use of robots in work-relatedscenarios. RoboCup@Work utilizes proven ideas and concepts from RoboCup competitions totackle open research challenges in industrial and service robotics. With the introduction of this newevent, RoboCup opens up to communities researching both classical and innovative roboticsscenarios with very high relevance for the robotics industry.

Examples for the work-related scenarios targeted by RoboCup@Work include

●loading and/or unloading of containers with/of objects with the same or different size,●pickup or delivery of parts from/to structured storages and/or unstructured heaps,●operation of machines, including pressing buttons, opening/closing doors and drawers, andsimilar operations with underspecified or unknown kinematics,●flexible planning and dynamic scheduling of production processes involving multiple agents(humans, robots, and machines),●cooperative assembly of non-trivial objects, with other robots and/or humans,●cooperative collection of objects over spatially widely distributed areas, and●cooperative transportation of objects (robots with robots, robots with humans).

The RoboCup@Work scenarios target difficult, mostly unsolved problems in robotics, artificialintelligence, and advanced computer science, in particular in perception, path planning and motionplanning, mobile manipulation, planning and scheduling, learning and adaptivity, and probabilisticmodeling, to name just a few. Furthermore, RoboCup@Work scenarios may also addressproblems for which solutions require the use and integration of semantic web technology, RFIDtechnology, or advanced computational geometry.

Solutions to the problems posed by RoboCup@Work require sophisticated and innovativeapproaches and methods and their effective integration. The scenarios are defined such that theproblems are sufficiently general and independent of particular industrial applications, but alsosufficiently close to real application problems that the solutions can be adapted to particularapplication problems with reasonable effort.

A RoboCup@Work competition has only recently become a feasible idea for several reasons: Thearrival of new, small, and flexible robot systems for mobile manipulation allow more university-based research labs to perform research in the above-mentioned areas. Advances and a revivedinterest in the use of simulation technology in robotics enable research groups to perform seriousresearch without having a full set of costly robotics and automation equipment available.

The robotics and automation industry is recently shifting its attention towards robotics scenariosinvolving the integration of mobility and manipulation, larger-scale integration of service roboticsand industrial robotics, cohabitation of robots and humans, and cooperation of multiple robotsand/or humans. Last but not least, there is a huge interest by funding agencies and professionalsocieties in well-designed and professionally performed benchmarks for industry-relevant roboticstasks. RoboCup@Work is designed as an instrument to serve all these needs.

2. Organization of the League

2.1 League Committees


5

The following list of committees is implemented for RoboCup@Work.

2.1.1 Technical CommitteeThe Technical Committee (TC) is responsible for technical issues of the league, most notably thedefinition of the rules, the qualification of teams, the adherence to the rules as well as theresolution of any conflicts that may arise during tournaments. The current TC members are:

●Walter Nowak, Locomotec, Germany●Nico Hochgeschwender, Bonn-Rhein-Sieg University, Germany●Gerhard Kraetzschmar, Bonn-Rhein-Sieg University, Germany●Rainer Bischoff, KUKA Laboratories, Germany●Jan Carstensen, Leibniz University Hannover, Germany●Yiyan Wang, Singapore Polytechnic, Singapore●Arne Hitzmann, Ostfalia University of Applied Sciences, Germany●Frederik Hegger, Bonn-Rhein-Sieg University, Germany

2.1.2 Organizing CommitteeThe Organizing Committee (OC) is responsible for the practical implementation of tournaments,most notably for providing test arenas and any objects and facilities required to perform the varioustests, scheduling tests, assigning and managing referees, recording and publishing competitionresults, and any other management duties arising before, during, and after a tournament. Thecurrent OC members are:

●Walter Nowak, Locomotec, Germany●Joscha David Fossel, University Liverpool, UK●Arne Hitzmann, Ostfalia University of Applied Sciences, Germany●Frederik Hegger, Bonn-Rhein-Sieg University, Germany●Rene Pegoraro, Universidade Estadual Paulista, Brazil●Jan Carstensen, Leibniz University Hannover, Germany

2.1.3 Industrial Advisory BoardThe role of the Industrial Advisory Board is to ensure the industrial relevance of the tests and theoverall competition. It allows representatives from industry to voice interesting new problems,which may be included in existing or new test scenarios. Currently, the Industrial Advisory Board has no members yet.

2.1.4 Research Advisory BoardThe role of the Research Advisory Board is to ensure that the tests and scenarios are innovativeand relevant from a research point of view. The members will be recruited from academicinstitutions and research organizations.Currently, the Research Advisory Board has no members yet.

2. 2 League InfrastructureIn order to provide a forum for continuous discussions between teams and other stakeholders, theleague builds and maintains an infrastructure consisting of web site, mailing lists, and repositoriesfor documentation, software, and data. The infrastructure is complemented by a minimuminfrastructure to be built and maintained by teams, i.e. teams should eventually create their ownweb page to which the RoboCup@Work League’s web pages can be linked.


6

2.2.1 Infrastructure Maintained by the LeagueThe official website of RoboCup@Work is at

http://www . robocupatwork . org

This web site is the central place for information about the league. It contains general introductoryinformation plus links to all other infrastructure components, such as a league wiki, the mailinglists, important documents such as this rule book, announcements of upcoming events as well aspast events and participating teams.

The league maintains several mailings lists:●rc-work@ lists . robocup . orgThis is the general RoboCup@Work mailing list. Anyone can subscribe, but a real name mustbe provided either as part of the emai address or being specified on the mailing list subscriptionpage. The list is moderated in order to avoid abuse by spammers. ●rc-work- tc @ lists . robocup . orgThis is the mailing list for the Technical Committee. Posts from non-members have to beapproved by the list moderator. Approvals will be given only in well-justified cases.

2.2.2 Infrastructure Maintained by TeamsEach team is requested to build and maintain a minimum infrastructure for the team. Thisinfrastructure consist of

●a team web site,●a team contact, and●a team mailing address.

The team web site should contain the following information:●name of the team, and team logo, if any●affiliation of the team●team leader including full contact information●list of team members●description of the team’s research interest and background●description of specific approach pursued by the team●description of the robot(s) used by the team●list of relevant publications by team members

The team contact should be the official contact of the team. Usually, for university-based teams,this would be an academic person such as a professor or post-doc, who should, however, beresponsive and be able to answer quickly when contacted by email.

The team mailing address should be an email alias, which should be used to subscribe the team tothe general RoboCup@Work mailing list. The email alias should at least include the team contactand the team leader.

3. Participation in the CompetitionParticipation in RoboCup@Work requires successfully passing a qualification procedure. Thisprocedure is to ensure the quality of the competition event and the safety of participants.Depending on constraints imposed by a particular site or the number of teams interested toparticipate, it may not be possible to admit all interested teams to the competition.


mailto:[email protected]














http://www.robocupatwork.org/





7

3.1 Steps to ParticipateAll teams that intend to participate at the competition have to perform the following steps:

1. Preregistration (may be optional; currently by sending email to the TC)2. Submission of qualification material, including a team description paper and possibly

additional material like videos or drawings3. Final registration (qualified teams only)

All dates and concrete procedures will be communicated in due time in advance.

3.2 QualificationThe qualification process serves a dual purpose: It should allow the Technical Committee to assessthe safety of the robots a team intents to bring to a competition, and it should allow to rank teamsaccording to a set of evaluation criteria in order to select the most promising teams for acompetition, if not all interested teams can be permitted. The TC will select the qualified teamsaccording to the qualification material provided by the teams. The evaluation criteria will include:

●Team description paper●Relevant scientific contribution/publications●Professional quality of robot and software●Novelty of approach●Relevance to industry●Performance in previous competitions●Contribution to RoboCup@Work league, e.g. by

○ Organization of events○ Provision and sharing of knowledge

●Team web site

3.3 Team Description PaperThe Team Description Paper (TDP) is a central element of the qualification process and has to beprovided by each team as part of the qualification process. All TDPs will be included in the CDproceedings of the RoboCup Symposium. The TDP should at least contain the followinginformation in the author/title section of the paper:

●Name of the team (title)●Team members (authors), including the team leader●Link to the team web site●Contact information

The body of the TDP should contain information on the following:●focus of research/research interest●description of the hardware, including an image of the robot(s)●description of the software, esp. the functional and software architectures●innovative technology (if any)●reusability of the system or parts thereof●applicability and relevance to industrial tasks

The team description paper should cover in detail the technical and scientific approach,while the team web site should be designed for a broader audience. Both the web site and the TDPhave to be written in English.


8

4. Organization of the Competition

4.1 TeamsThe Technical Committee and the Organizational Committee will jointly determine the number ofteams permitted to participate in a competition well in advance. The rules shall enable acompetition with up to at least 24 teams lasting not more than four full days.The number of people to register per team is not restricted by default, but may be limited due tolocal arrangements. Teams that plan to bring more than four members are advised to contact theOrganizing Committee beforehand. During registration, each team has to designate one member as team leader. A change of the teamleader must be communicated to the Organizing Committee. The team leader is the only personwho can officially communicate with the referees during a run, e.g. to decide to abort a run, to calla restart, etc. The team leader can ask the OC to accept additional teams members for thesetasks.

During on-site registration and upon request by the Organizing Committee a team has to nominateone or more referees for the competition. If a team fails to provide referees in an appropriate way,the Organizing Committee may take actions such as reduction of points for the respective team.

4.2 Team Practice and Use of ArenasThe teams will be given an opportunity to practice with their robots either in the competition arenasor in special test arenas, if available. The frequency and lengths of practice periods will be decidedby the OC on site. The OC will also decide about if and how many teams may use an arenasimultaneously and can decide on a practice schedule for teams wishing to use the arenas. Arenasmay be modified between practice time and competition runs.

4.3 Stages and TestsThe OC may decide to split the competition into several stages. The competition design mayforesee that only a smaller number of teams qualifies for a consecutive stage. An exemplarycompetition design could foresee a first stage with all qualified teams, a second stage with only thebest 10 teams from the first stage, and a finals stage with the best 5 teams of the second stage.

Each stage is composed of a sequence of tests. The OC and the TC will jointly determine the typeand number of tests in a stage and schedule the tests. Each test may be executed in one or multiple runs. The term run designates a single trial of a testfor each team.

4.4 Scoring and RankingFor each test the calculation of scores is defined individually, comprising points for achievingcertain subtasks, points for winning a run, negative points for reducing the difficulty, and penaltypoints.

If not specified otherwise, the following set of scoring rules applies for each test:

Teams may use simplifications, which will result in a reduction of scores for the given run:


9

● Use of external sensors −200 points● Use of other external objects (e.g. to support localization): −100 points● Use of own loading or unloading areas: −200 points

Additional simplifications are specified for individual tests. These reductions do not count aspenalty points. Teams that want to make use of the simplifications above have to announce them inadvance of the competition to the TC. The TC might forbid use of specific elements forsimplification if these are not in the spirit of the league or may cause disproportionate advantagesfor a team.

Penalty points are given as follows, each time again the incident occurs:

● A manipulation object is lost or placed anywhere outside of the service areas: −100 points.● Robot caused a collision with the environment: −50 points.● Restart used: -200 points.

Controlled collisions between a robot and the environment are allowed, i.e. slightly touching aloading area with the robot’s base or gripper. As soon as a robot starts to move environmentalelements or the wheels start to spin (wheel-spin), it will be considered as a collision and therunning test will be aborted. A controlled collision can only occur during manipulation operations.

If the projection of any part of a robot touches a barrier tape marking a virtual wall it counts as acollision. Touching or Passing a Barrier Tape when entering or exiting the arena does not count ascollision.

When an object lost contact to the robot or touches the ground, it is not considered as beingcarried anymore (e.g. an object falls off the gripper or from the transportation platform).

All teams fully complete a perfect run will receive a completion bonus of 0.75 points per second lefton the run time. These points are only awarded if the run is perfect, all objectives reached withoutany penalties.

A team cannot get less than zero points for one run. The score for a test is the average of the runsof this test, calculated by summing up the scores of each run and dividing by the number of runs.The scores of the tests of the first stage are summed up, and the teams with the highest sumsproceed to the next stage.

In case of a tie, the team with the higher score during the last test wins. If both had the samescore, then the test before that one is considered and so on. In case of a tie, there will be a deciderrun chosen by the TC only for those teams affected.

Some tests have three so-called complexity levels. A complexity level (low, medium and high)encodes the difficulty of the test such as the number of objects to manipulate or whether the arenais equipped with obstacles or without.

Every team decides for each run of a test in which complexity level they want to participate. Thisdecision must be made by all teams when asked for by the OC, but at least before the first teamstarts a particular run.


10

4.5 Common Procedures The order in which the teams have to perform a run is determined by a draw by the OrganizingCommittee. The order will be made public at least one hour before slot of the particular test. A runis preceded with a 5 minutes preparation time. This time begins once the previous team has leftthe arena and the arena is setup (e.g. all manipulation objects are in place).

During the preparation time the team may enter the competition area in person, but not with therobot. Entering before this time is allowed only with explicit allowance of the OrganizingCommittee.

When the preparation time ends, all team members must immediately leave the area androbots must be ready to start. The team has to check if the arena is setup correct (e.g. allmanipulation objects are placed according to task spec, obstacles are placed correct).

The referees start a run by sending the start signal from the Referee Box.

A run ends when the duration for the given test has passed, when the task has been finished by allrobots, when the referees decide to stop it, or when the team leader of the team whose turn it isdecides that the run can be finished earlier as no more progress is expected.

During a run teams may only interact with the robot or enter the area if explicitly allowed by thereferees.

After each run, the teams must leave the arena within one minute.

4.5.1 RestartsTeams might use one so-called restart in a run. Restarts have the following aspects:

● Per run, at most one restart is allowed for a team, if not specified otherwise in a test.● At any time during a run, the team can call for a restart to the referees.● When the referees accept the call for restart, the team may enter the arena. The time will

continue running.● The arena and the robot will be reset exactly to the setup at the beginning of the run

(except the timer for the run). Random elements such as obstacles or object positionsremain like before.

● The points for this run achieved so far are reset to zero (including possible penalty points).After the run, a penalty score for the restart will be applied.

● The referees decide when the arena is prepared again for the restart. If the robot is not yetready, teams can keep trying to get it ready until the time for the run is over.

● As soon as the team signals that the robot is ready, the task specification is sent again.● Afterwards the start signal is sent from the referee box.

4.6 Roles of OC and TC


11

The Organizing Committee is responsible for all aspects concerning the competition, such asproviding the competition arenas, ensuring their conformity with the rules, assigning space toteams in the team area, the organization and scheduling of meetings, the nomination andscheduling of referees, and the scheduling and timely execution of tests and stages. The Technical Committee is responsible for all aspects concerning the rules, such as complianceof the robots with rules and safety standards, the clarification of rules and resolution of anyconflicts that may arise during the competition.

4.7 RefereesThe referees are to ensure the correct execution of the tests. They may interrupt runs if theysuspect breaches of rules, see possible danger for humans or possible damages of robots or theenvironment. The referees are also responsible for timekeeping and recording scoring results.

4.8 Meetings and Language of CommunicationBoth the TC and the OC may organize several special meetings during a competition, such asreferee meetings, team leader meetings, etc. The meetings will be announced locally. It is theresponsibility of the team to inform itself about the organization and scheduling of such meetings.

Each team is expected to send at least one representative to such meetings. If the meeting refersto specific roles, such as “referee” or “team leader”, the person designated by the team to fill thisrole is expected to participate.

The language for all communication in the league is English.

4.9 Code of Conduct and DisqualificationTeams and team members are expected to maintain a friendly and cooperative atmospherethroughout a competition and contribute to a vivid work environment and to scientific exchangebefore, during and after a competition.

The Technical Committee may disqualify individual team members or a whole teams during acompetition for severe reasons, such as repeated breach of rules.

5. General Design of RoboCup@Work ScenariosEach of the particular tests defined later in this document may define its own scenario. In thisdocument, a scenario consists of the following elements:

● the environment● the objects that affect navigation● the objects that are to be manipulated● the objects with which robots interact● the number of robots allowed per team● the number of teams competing simultaneously in the same arena● the task to be performed by a team● the criteria for evaluating a team’s performance

In order to avoid excessive development efforts for each specific test and to allow reuse of partialfunctionalities the scenarios are built from a reasonably small set of components, which are laterput together in different ways. This section describes these elements.


12

5.1 Assumptions about Robots used in the CompetitionThe following assumptions are made about the kind of robots used in the competition:

●At least one of the robots used by a team is mobile and moves on wheels. No specificassumptions are made about the kinematic design, but the mobile robots should be able tomove on basically flat, sufficiently firm surfaces. ●The robots have at least one manipulator and are able to grasp objects, which are graspableby a parallel gripper with a jaw width of at least 5 cm and do not weigh more than 300 g.●The robots use sensors to obtain information about their whereabouts in the environment andthe task-relevant objects. The major types of sensors that may be used by the robots include:

○laser range finders (cf. models by Hokuyo or Sick)○color CCD cameras (cf. any kind of USB camera)○3D cameras (such as the Kinect camera)

The design of the scenario should be such that the robots can solve the tasks safely androbustly using (all or a subset of) these sensors.

●Future competitions may foresee the use of RFID sensors in the scenario design.

5.2 Rules Applying to RobotsThe robots used for competition shall satisfy professional quality standards. The concrete definitionof these standards is to be assessed by the Technical Committee, comprising aspects such assturdy construction, general safety, and robust operation. It is not required that the robots arecertified for industrial use.

5.2.1 Robot Design and Design ConstraintsThe robots need to comply with certain size constraints. A robot, including all parts attached to it asused in the competition, must be able to move by itself into a configuration so that it fits into a cubeof side lengths 80 cm x 50 cm x 80 cm (length x width x height). If all the robot’s parts, such asmanipulator or anything able to protrude outside of the previously specified cube, are fullyextended, the system must still not exceed a cube of side lengths 120 cm x 80 cm x 160 cm(length x width x height). The organizers may specify further constraints, such as weight limits.

The manipulator of the robot should be designed and mounted on the robot such that it can graspobjects which will be placed in heights of between 0 cm and 40 cm above the floor.

Electric, pneumatic, and hydraulic actuation mechanisms are permitted, provided that they areconstructed and produced according to professional standards and meet safety constraints.Combustion engines and any kind of explosives are strictly forbidden. Robots may not pollute orharm their environment in any way, e.g. by loss of chemicals or oil, spilling liquids, or exhaustinggases. Furthermore, constraints on the noise generated by a robot in operation may apply. Thesewill be communicated in due time.

If there are even vague doubts about the eligibility of using particular designs, parts, ormechanisms, the team should consult the TC well in advance.

The robots have to be marked such that a clear distinction of robots used by different teams duringa test is possible for spectators. The Organizing Committee can define the concrete types ofmarkers to be used. In this case the markers are not taken into account when checking the robot’ssize constraints. The markers shall not interfere with safe operation of the robot.


13

The TC may require that robots are equipped with a wireless communication device of some sort(e.g. 802.11n), in order to communicate task specifications to the robots.

5.2.2 Robot Behavior and SafetyIn general, all robots shall be operated with maximum safety in mind. Any robot operation must besuch that a robot neither harms humans nor damages the environment. A team must choose theoperating parameters of their robot, e.g. the motion velocities for a robot base or a manipulator, orthe grasp forces of a gripper, such that it can guarantee the safe operation of their robots.

All robots must have a mechanical mechanism for immediately stopping the robot in cases ofemergency. This mechanism must be clearly visible and easily accessible. The OrganizingCommittee may request the proof of a robot’s safety (e.g. the correct operation of an emergencystop) anytime during the tournament and exclude teams that cannot satisfy safety requirements.

When participating in a tournament, the team may operate the robot only in their own team area, inthe arenas provided (possibly constrained by a schedule assigning periods of time for exclusiveuse of the arena by a team or a group of teams), and in any other areas designated by theorganizers for robot operation. Any operation of robots outside of these areas, e.g. in public areasor emergency paths, require prior permission by the Organizing Committee.

5.3 Use of External DevicesNo external devices are allowed (e.g. remote controls) in general. Exceptions may be certainsimplifications leading to reduction of points as described in Chapter 4.4, or in particular tests. Allcommunication of the robots with external elements must be wireless. Cable connections betweenthe robot and external devices are not allowed during competition runs.

The TC shall provide a referee box1, which communicates task specifications and starting signalsto the robots. It is possible for the TC to choose an alternative referee box software, or to allowteams to use their own referee box. This must be announced before the tournament starts. A team may set up an external computer to monitor the operation of the robot(s) during a run. Thismonitoring system must be designed such that no manual interaction through keyboard, mouse, orany other input device is required during a run. Team members must keep their hands off thekeyboards and mice of all their computers during a run.It must be clear at all times that no manual or remote control is exerted to influence the behavior ofthe robots during a run. Exceptions may be specified by particular tests, e.g. for tasks wherehanding over objects to humans is required.

5.4 Design of Competition Arenas

5.4.1 Size of the ArenaThe size of a competition arena is a rectangular area no less than 2 m x 4 m and no more than 10m x 12 m.An orientation is always associated with the arena. An arbitrary wall is designated as “North”orientation, and the wall to its right is designated as “East” and so on. The orientations will beassigned by the local league chair or/and the TC as soon as the arena is built up. Figure 5.1 showsone possible example of an arena configuration.1 https :// github . com / robocup - at - work / referee - box


https://github.com/robocup-at-work/referee-box















14

Figure 5.1: An exemplary setup for a RoboCup@Work environment

5.4.2 FloorThe floor is made of some firm material. Examples include floors made of concrete, screed, timber,plywood, chipboard, laminated boards, linoleum, PVC flooring, or carpet. Some examples areillustrated in Figure 6.1. Floors may neither be made of loose material of any kind (gravel, sand, orany material which may damage the functioning of the robots’ wheels) nor may such material beused on top of the floor. Liquids of any kind are not allowed. The floor may have spots ofunevenness of up to 1cm in any direction (clefts, rifts, ridges, etc.).

Figure 5.2: Examples of floors that can be used for RoboCup@Work arenas.

5.4.3 WallsThe competition arena is partly surrounded by walls. The height of the walls is no less than 20 cmand no more than 40 cm. One or more gates may be foreseen, where robots can enter or leave thearena. Gates may or may not be closable. The walls have a mostly uniform color.

5.4.4 Service Areas


15

Arenas contain one or more service areas, which have specific purposes for a particular test.Examples include loading and unloading areas, conveyor belts, rotators, storage areas, etc.Service areas may contain specific environment objects, such as racks, shelves, etc.

5.4.5 RacksService areas, e.g. loading and unloading areas, may foresee the use of racks. Objects to bedelivered or removed from racks have to be placed or picked up from the top. The height of theracks should be no less than 5 cm and no more than 40 cm. The color for the top surface of racksis white, unless a test specifies a different color. The top surface of the rack may be speciallydesigned in order to serve specific purposes, e.g. holding objects.

5.4.6 ShelvesService areas may foresee the use of shelves and shelf units. Objects to be delivered or removedfrom shelves have to be placed or picked sideways. The height of the shelves should be no lessthan 5 cm and no more than 40 cm. The color of the shelf surface is white, unless a test specifies adifferent color. The shelf surface may be specially designed in order to serve specific purposes,e.g. holding objects.

5.4.7 PlacesAn arena designed for a particular test may foresee the definition of a set of designated places,which are locations in the arena that can be referred to by a unique, symbolic identifier. Theseidentifiers are used e.g. for task specification, possibly in conjunction with other information, suchas an orientation. Places may be marked by markers of some sort.

5.4.8 ObstaclesAn arena defined for a particular test may foresee the use of obstacles. Obstacles may by passive(i.e. not able to relocate by themselves) or active (e.g. other robots). The size of obstacles shouldbe no less than 10 cm x 10 cm x 5 cm; there is no upper bound on the size. The details about thedynamic objects are not known before the competition and will be chosen by the OC onsite.Examples for obstacles are trash bins, boxes, big aluminium profiles or even other robots.

5.4.9 Floor MarkersThe arena used for a particular test may foresee the use of floor markers for designated places.The design of these floor markers are rectangular black-and-white images as used by theARToolKit library:

● http :// www . hitl . washington . edu / artoolkit /

The black inner square of the markers is at least 8 cm x 8 cm large with an additional white borderaround of at least 12 cm side length. Figure 5.4 depicts some examples of these markers.


http://www.hitl.washington.edu/artoolkit/












16

Figure 5.3: Example of a floor marker.

5.4.10 Barrier Tape as Virtual WallsThe arena may include virtual walls marked by either striped yellow and black or white and red barrier tape on the floor (see Figure 5.3.1). If any part of a robot passes over such a tape it is considered like a collision with a normal wall.

Figure 5.3.1: Example of Barrier Tape.

5.4.11 Entrance and ExitIf possible the Arena should have one entrance and one exit, which are equipped with Laser barriers. These barriers are used for timekeeping and will be used by the Referee Box.


17

5.4.12 Setup of the ArenaThe arena must be surrounded by either walls or barrier tape. The only exceptions are the Entrance and Exit. There must be a route to all places and markers relevant for a test which has a minimum width of 55 cm..

5.5 Design of Manipulation Tasks

5.5.1 Manipulation ObjectsThe manipulation objects in RoboCup@Work shall include a wide range of objects relevant inindustrial applications of robotics and eventually cover any raw material, semi-finished parts, andfinished parts and products as well as tools and possibly operating materials required formanufacturing processes.

The intention is to start with a simple set of objects of different shapes and colors and to widen thespectrum every year in at least one aspect. The initial set of objects includes basic and standardscrews and nuts in various sizes and weight as shown in the following table:

SymbolicDescription

Weight (in g)

Details

F20_20_B 49gHeight: 20mmWidth: 20mm

Length: 100mm

F20_20_G 49gHeight: 20mmWidth: 20mm

Length: 100mm

S40_40_B 186gHeight: 40mmWidth: 40mm

Length: 100mm

S40_40_G 186gHeight: 40mmWidth: 40mm

Length: 100mm


18

M20_100 296gISO 4014

M20Length: 100mm

M20 56g ISO 4032 M20

M30 217g ISO 4032M30

R20 14gInner diameter: 20mmOuter diameter: 30mm

Length: 45mm

V20 14gInner diameter: 20mmOuter diameter: 30mm

Length: 45mm

Table 5.1: Examples of typical manipulation objects

Objects of one kind can slightly vary e.g. considering surface. Teams can ask the TechnicalCommittee for allowance to use objects brought by the team to a competition, provided they do notdiffer in a relevant way from the default objects.

5.5.2 Placement of objects

For the placement of manipulation objects the following terms are used:● position: point within 2D coordinate system of service area● rotation: rotation around vertical axis of service area● orientation: rotation around horizontal axes of service area, i.e. whether the object is

standing upright or lying on its side.● pose: combination of position and rotation


19

5.5.3 Grasping objectsIf not specified differently in a test, the following definition applies to decide if an object counts as grasped from a service area.

An object counts as grasped from a service area, when1. the robot has lifted the object. It is not enough to shove it beyond the service areas edge.2. the object was moved outside of the source service area. Outside means that the vertical

projection of the object’s convex hull does not touch the service area anymore.3. the robot shows the intention to move itself with the object away from the service area. By

default this is assumed when the distance on the horizontal plane between every part of therobot and service area as well as between every part of the object and service area is morethan 20 cm; or when the object has been placed on another service area. The referees mayinterpret the intention differently if agreed on before the run.

The last point shall enable to let the robot pick up an object in order to analyse its type, e.g. by holding it close to a camera on the robot.

If the robot handles an object, but does not fulfil all points above, the object does not count as grasped, and neither points for grasping a required object, nor penalty points for grasping an unspecified object are given. Still, if the object drops to the ground or an uncontrolled collision occurs, the normal penalty points apply.

5.5.4 Placing objects on service areasIf not specified differently in a test, a manipulation object counts as placed on a target service area if any part of the object is touching the surface of the service area and the object is not moving at the end of the run.

The position, rotation or orientation of the object on the service area can be chosen freely by the robot.

5.6 Overall Scenario and Variation PointsThe following two figures illustrate the kind of arenas that can be built with the previously describedcomponents. Figure 5.4 shows an old arena configuration (from IROS 2012) as a topological mapdefining places and connections, while Figure 5.5 shows an arena configuration (from GermanOpen 2013) with only a few surrounding walls, more open space and a conveyor belt.


20

Figure 5.4: Example of an arena with 5 service areas (D1, D2, S1, S2, S3). The purple squares define places.

Figure 5.5: Illustration of the arena configuration used at the German Open 2013.


21

6. TestsThe actual competition tasks are defined as a set of so-called tests. A competition is organized inseveral stages. For each stage, the Technical Committee will select a set of tests to be performedby each team.

A possible stage design would be as follows:

Stage 1:●BNT: Basic Navigation Test●BMT: Basic Manipulation Test●BTT: Basic Transportation Test

Stage 2:●CBT: Conveyor Belt Test●PPT: Precision Placement Test

Finals: ●A combination of tests from previous stages

Note that not necessarily the most complex tests have to be in the finals; it might make more senseto use them as early-bird benchmark challenges for things most teams are not yet able to solve.Tests for the finals should be tests which can be solved rather well by all finalists, but with differentperformance, in order to increase the attractiveness of the event. The ability of a team to solvemore complex tests should be reflected by the prior score the team takes into the finals.


22

6.1 Basic Navigation Test (BNT)

6.1.1 Purpose and Focus of the TestThe purpose of the Basic Navigation Test is to check whether the robots can navigate well in theirenvironment, i.e. in a goal-oriented, autonomous, robust, and safe way.

As the navigation problem is in the focus of robotics research for a long time and many approachesfor solving it and its subtasks (like exploration, mapping, self-localization, path planning, motioncontrol, and obstacle avoidance) exist, the focus of this test is to demonstrate that theseapproaches function properly on the robots used by the teams and in the environment defined bythe test.

6.1.2 Scenario EnvironmentThe arena used for this test contains all elements that affect or support navigation: walls, serviceareas, places, arena objects, wall markers, and floor markers. In addition, obstacles may be placedin the environment.

A possible arena defined for this test is displayed in Figure 5.4 and 5.5.

6.1.3 Manipulation ObjectsThis test does not include any objects for manipulation.

6.1.4 TaskFor the navigation test, a single robot is used. The robot will be sent a task description, which is astring containing a series of triples, each of which specifies a place, an orientation, and pauseduration. The robot has to move to the places specified in the task string, in the order as specifiedby the string, orient itself according to the orientation given, cover a place marker, pause itsmovement for the time in seconds as specified by the pause length, and finally leave the arenathrough the gate.

The task specification consists of: ● A destination location, e.g. S1, D2, T7 or U4● An orientation (N, S, W, E)● A duration in seconds

Two example task specifications for the BNT test are given below:● BNT<(S6,N,3)(S2,N,3)(D1,S,3)(S5,W,3)(D3,E,3)(D4,S,3)>● BNT<(S6,N,3)(S3,W,3)(S7,W,3)(S2,W,3)(D3,E,3)>

The duration is always set to 3 seconds in order to make validation easier for the referees.

6.1.5 Complexity Options


23

● Obstacle complexity (pick one):○ Easy obstacles (bonus factor = +0.2):

■ there are up to two obstacles in the arena○ Medium obstacles (bonus factor = +0.4):

■ there are up to three obstacles in the arena■ the placement of the obstacles is harder

● Barrier Tape complexity (bonus factor = +0.4):○ there are up to two barrier tape obstacles in the arena

● Navigation complexity (pick one):○ Easy navigation (bonus factor = + 0.0):

■ the place marker has to be covered in such a way by the robot that at least apart of the black area covered

○ Medium navigation (bonus factor = + 0.1):■ the place marker has to be covered in such a way by the robot that at least a

small part of the black area is covered ■ the orientation must be correct, i.e. the robot must not deviate more than 45

deg.○ Hard navigation (bonus factor = + 0.2):

■ the place marker has to be fully covered by the robot ■ the orientation must be correct, i.e. the robot must not deviate more than 45

deg.

6.1.6 RulesThe following rules have to be obeyed:

●The order in which the teams have to perform will be determined by a draw.●A team has a time period of 5 minutes to complete the task, in addition to the defaultpreparation time.●At the beginning of a team’s run, the team will get the task specification, which is a sequenceof triplets (<place>, <orientation>, <break>), where <place> designates one of the specifiedplaces of the environment (e.g. D0, D1, D2, O1, S2, S3, T2, T4, T6), <orientation> is one of(N,E,S,W), and <break> is a one-digit integer between 1 and 3. An example task specificationwould be <(S1,E,1),(D1,W,1),(S3,E,1),(D2,W,1)> . ●After the team’s robot enters the arena, it must move to the places given in the taskspecification and assume the orientation specified after the place. The robot may reach adestination by choosing any path.●The robot must visit the places in the order given by the task specification. It is possible toskip a place of the task specification and continue with the next one. In cases where the robotskipped one or multiple places there may be multiple possible matchings between placesreached and places specified. In that case for calculating scores the matching is taken whichleads to the highest score for the team.●A destination is counted as reached when the robot covers the place marker as much as thecomplexity level demands. The orientation must not deviate more than 45 deg.●When a destination is reached, the robot must stop movement for the number of secondsspecified by the break. ●The time is stopped when the robot has completed the task and left the arena. If the teamcannot complete the task within 5 minutes, the run will be stopped after 5 minutes.


24

6.1.7 Scoring●The team will receive 50 points for reaching a destination correctly (place and orientation) asgiven in the task specification and provided it stops for the time specified. ●The team receives a penalty of –50 points each time the robot touches an obstacle, a wall, anarena object or a service area (i.e. any contact with the environment).●50 points are awarded for completing the task specification completely correct, i.e. visiting alldestinations from the task specification according to position and orientation (according to thechosen complexity level) and finally leaving the arena. ●The reached points of a test will be multiplied with the complexity factor that belongs to thechosen complexity level.

6.2 Basic Manipulation Test (BMT)

6.2.1 Purpose and Focus of the TestThe purpose of the Basic Manipulation Test is to demonstrate basic manipulation capabilities bythe robots, like grasping, turning, or placing an object.

The focus is on the manipulation and on demonstrating safe and robust grasping and placing ofobjects of different size and shape. Therefore, only minimal movement of the robot is required.

Some minor movement is intentionally designed into this test in order to force the teams to dodynamic assessment of the situation (e.g. estimating positions of manipulation objects, determininggrasp positions, etc.) and to avoid that solutions depending on completely known initial situationsand well-calibrated systems are possible.

6.2.2 Scenario EnvironmentThe arena used for this test contains basically all elements as for the Basic Navigation Test.Additionally to environmental elements (walls, service areas, floor markers, etc.), differentmanipulatable objects will be placed on the service areas.

6.2.3 Manipulation ObjectsThe manipulation objects in this test include the objects specified in Section 5.5.1.

6.2.4 TaskA single robot is used. The robot can be placed in an arbitrary starting location by the team. Thetask consists of a sequence of grasp and place operations, with a small base movement inbetween. The objective is to move a set of objects from one service area into another. The task isfinished once all objects are moved or when the time foreseen for the run ends.

The task specification consists of: ● The specification of the initial place (e.g. D0, S5, U2)● A source location, given as place (any one)● A destination location, given as place (any one, but nearby the source location)● The specification of a final place for the robot (which does not need to be reached)

Manipulation objects will be referred to by a symbolic description, as defined in Section 5.5.1.


25

For now, in the competitions only the line is just as possible configuration:● Line( <sequence of objects> )

Further configurations may be specified in the future. Note, that the pose and orientation of eachobject in the destination service area is left unspecified and can be chosen by the robot. Thatmeans also objects don’t have to be put in a line.

Two examples for a full task specification is as follows:● BMT<S6,S6,S7,line(V20,R20,F20_20_B),S7>● BMT<S6,S6,S7,line(F20_20_G,F20_20_B,M20_100),S7>

6.2.5 Complexity Options

● Manipulation Object complexity (pick one):○ Choose objects (bonus factor = 0.0):

■ The team can freely choose the manipulation objects based on the setspecified in Section 5.5.1.

○ Few objects (bonus factor = 0.2):■ The team can freely choose a set of five manipulation objects based on the

set specified in Section 5.5.1.○ All objects (bonus factor = 0.5):

■ All manipulations objects based on the set specified in Section 5.5.1 can beused.

● Decoy Object complexity (pick one):○ No decoy objects (bonus factor = 0.0):

■ No decoy objects will be used.○ Few decoy objects (bonus factor = 0.2):

■ The team can freely choose a set of three manipulation objects based on theset specified in Section 5.5.1, to be used as decoy.

○ All decoy objects (bonus factor = 0.3):■ All manipulations objects based on the set specified in Section 5.5.1 can be

used as decoy.

● Orientation Complexity (bonus factor = 0.2):○ The manipulation objects can be placed in all orientations.

● Rotation Complexity (bonus factor = 0.2):○ The manipulation objects can be placed in all orientations.

● Position Complexity (bonus factor = 0.2):○ The manipulation objects while be placed by the referees.


●The order in which the teams have to perform will be determined by a draw.●A team has a time period of 5 minutes to complete a run.●At the beginning of a team’s period, the team will get the task specification. ●The team can setup the robot anywhere inside the arena.●A manipulation object counts as successfully grasped as defined in Section 5.5.3.


26

●A manipulation object counts as successfully placed, if the robot has placed the object into thecorrect destination service area as described in Section 5.5.4.. ●The team has at most 5 minutes to complete the task. The time is stopped when the robot hascompleted task by placing the last object of the task specification. If a team cannot completethe task within 5 minutes, the run will be stopped after 5 minutes.

6.2.7 ScoringPoints are awarded as follows:

● 100 points are awarded for successfully grasping a manipulation object that is part of thetask specification..

● 50 points are awarded for successfully placing a manipulation object into the destinationservice area.

● - 100 points for grasping a wrong manipulation object.● 50 points are awarded for completing the task specification completely correct. ● The reached points of a test will be multiplied with a defined complexity factor depending on

the previously chosen complexity level.


27

6.3 Basic Transportation Test (BTT)

6.3.1 Purpose and Focus of the TestThe purpose of the Basic Transportation Test is to assess the ability of the robots for combinednavigation and manipulation tasks. The focus of the Basic Transportation Test is to assess the ability of a team to deal with flexibletask specifications, esp. concerning information about object constellations in the source and thetarget locations, and task constraints such as limits on the number of objects allowed to carrysimultaneously, etc.

6.3.2 Scenario EnvironmentThe arena used for this test contains all elements as for the Basic Manipulation Test.

6.3.3 Manipulation Objects

The manipulation objects in this test include the objects specified in Section 5.5.1.

6.3.4 TaskA single robot is used, which is initially positioned outside of the arena near a gate to the arena.The task is to get several objects from the source service areas (such as S1, T7, or U3), and todeliver them to the destination service areas (e.g. D1 and D3). Robots may carry up to threeobjects simultaneously.

The task specification consists of two lists:● The first list contains for each service area a list of manipulation object descriptions. The

descriptions are similar as those used for the Basic Manipulation Test. ● The second list contains for each destination service area a configuration of manipulation

objects the robot is supposed to achieve. The configuration specification is similar as usedin the Basic Manipulation Test.

Two example task specifications are given here:● BTT<initialsituation(<S6,line(M20_100,F20_20_B)><S7,line(V20)>);goalsituation(<S1,line(

M20_100,F20_20_B)><S2,line(V20)>)>● BTT<initialsituation(<S6,line(S40_40_B,F20_20_B)><S7,line(M20_100)>);goalsituation(<S

1,line(F20_20_B,F20_20_B)><S2,line(M20_100)>)>

The term “line” in the task specification can be ignored.

6.3.5 Complexity OptionsThe same complexity scoring as in BMT applies.


●The order in which the teams have to perform will be determined by a draw.●A team has a time period of 10 minutes to complete a run.●At the beginning of a team’s period, the team will get the task specification. The robot has tostart from outside the arena.


28

●After the team’s robot starts, it must move into the arena and attempt to complete the task. ●A manipulation object counts as successfully grasped as specified in Section 5.5.3. ●It is not allowed to place manipulation objects anywhere except for the robot itself and any ofthe available service areas.●A robot may carry up to three objects at the same time.●The time is stopped when the robot has completed the task (delivered all objects to the rightlocations and left the arena through the exit gates). If a team cannot complete the task within10 minutes, the run will be stopped after 10 minutes. ●The team must leave the arena within a minute after completing the task.


● 75 points are awarded for successfully grasping a manipulation object required in the taskspecification..

● - 75 points if a wrong object has been grasped● 75 points are awarded for successfully placing a manipulation object into the destination

service area.● 50 points are awarded for completing the task specification completely correct. ● The reached points of a test will be multiplied with a defined complexity factor depending on

the previously chosen complexity level

6.4 Conveyor Belt Test (CBT)

6.4.1 Purpose and Focus of the TestThe purpose of the Conveyor Belt Test is to assess the robot’s ability to manipulate moving objectswhich are placed on a conveyor belt. The test demands fast perception and manipulation skills inorder to pick up objects from an operating conveyor belt.

6.4.2 Scenario EnvironmentThe same arena as for the Basic Manipulation Test is used (Figures 5.4 and 5.5). In case that thearena does not already include a conveyor belt (see Figure 6.1), it will be added only for thisparticular test.The OC may choose to use a rotating turntable instead of a linear conveyor belt in order to simplifythe setup of the arena. The particular choice shall be announced before the tournament.An area besides the conveyor belt is left free so that a robot can drive there and grasp the objectsfrom the belt. The other sides might not be accessible, because the device is placed into a corner. The conveyor belt should have a wireless mechanism to turn it on.


29

Figure 6.1: Illustration of a conveyor belt used in the competition

6.4.3 Manipulation ObjectsThe same manipulation objects as for the Basic Manipulation Test are used. The team has tochoose five objects out of the total object set (see Section 5.5.1). From those five objects, the TC isgoing to select three objects which will be placed on the conveyor belt. Beside these three objectsthere will be no other objects on the conveyor belt.

6.4.4 TaskA single robot is used. The robot is placed by the team in some starting position outside of thearena. Initially the conveyor is switched off (i.e. the belt is not moving) and all objects are put onthe belt with a distance between them determined by the TC. The task of the robot is to navigate tothe location of the conveyor belt and to grasp all objects from the moving belt before they fall off atthe other end of the belt. If a turntable is used, the objects can pass multiple times in front of therobot, until the maximum time for the run is over. The robot is supposed to place the graspedobjects on the robot itself. The task is finished once the robot successfully grasped all objects orthe foreseen time of the test has been exceeded.

There are two options to determine when to switch on the conveyor belt:● The robot starts it itself, using the wireless mechanism.● A team member can give a signal to the referees to switch on the conveyor belt.

Teams can decide during the run to give the signal to the referees, e.g. when the robot failed tostart it itself.

The task specification has the following structure:

CBT<srv_area>

where ‘srv_area’ is the location of the conveyor belt (e.g. T1)

6.4.5 Complexity OptionsAll Complexity Options from BMT apply.

● Speed Complexity (pick one): ○ low speed (bonus factor = +0.0):


30

■ the conveyor belt speed will be not more than 0.5 cm/s○ medium speed (bonus factor = +02):

■ the conveyor belt speed will be not more than 0.75 cm/s○ high speed (bonus factor = +0.4):

■ the conveyor belt speed will be not more than 0.10 cm/s


● The order in which the teams have to perform will be determined by a draw.● A team has a time period of 10 minutes to complete a run. The time is stopped when the

robot has completed the task.● At the beginning of a team’s period, the team will get the task specification.● The objects are placed on the conveyor before the run starts, for low complexity by the

team or for medium and high by a referee..● After the team’s robot starts, it must move inside the arena.● Either the robot itself starts the conveyor belt, or a team member can give a signal to the

referees to switch on the conveyor belt..● The objects have to be grasped actively from the moving belt. The robot is not allowed e.g.

to wait at the end with a particular basket and collect the falling objects from the end of theconveyor belt. Further, wiping the objects to the side is also not allowed.

● A manipulation object counts as successfully grasped, if the robot has lifted the object andmoved it outside of the area of the conveyor belt as specified in Section 5.5.3 point 2.

● The team must leave the arena within a minute after completing the task.


● 200 points are awarded for successfully grasping an object from the conveyor belt ● -150 points are given if an object dropped onto the ground, is placed not on the robot itself

or dropped from the end of the conveyor..● -100 points are given if the referees had to switch on the belt manually after signalled by a

team member, while a working automatic wireless mechanism to switch it on was available.● 50 points are awarded if the task has been fully achieved (when all objects placed on the

belt are grasped).


31

6.5 Precision Placement Test (PPT)

6.5.1 Purpose and Focus of the TestThe purpose of the Precision Placement Test is to assess the robot’s ability to grasp and placeobjects into object-specific cavities. This demands advanced perception abilities (to recognize thecorrect cavity for each object) and manipulation abilities (to grasp and place the object in such amanner that it fits into the cavity).

6.5.2 Scenario EnvironmentThe same arena as for the Basic Manipulation Test is used (Figures 5.4 and 5.5) whereas theplane of one service arena includes object-specific cavities as shown in the Figure 6.3 and Figure6.4. For each object used in the test there will be one specific cavity. The cavity has the dimensionof the object plus a 2 mm offset for each dimension. At most five cavities are used in the test.

Figure 6.3: Illustration of horizontal cavities for the different kind of objects

Figure 6.4: Illustration of vertical cavities for the different kind of objects

6.5.3 Manipulation ObjectsThe same manipulation objects as for the Basic Manipulation Test are used.

6.5.4 TaskA single robot is used. The robot is placed by the team freely within the arena. The objective of thetask is to pick the object which is placed in this service area and make a precise-placement in the


32

corresponding cavity at the service area with the special PPT platform (an example configuration isillustrated in Figure 6.5).

Figure 6.5: The PPT platform including five cavity tiles

The task consists of multiple grasp and place operations, possibly with base movement inbetween, which will, however, be short. The set of objects which are used in the test is known priori(at most the number of objects specified in Section 5.5.1). The task is finished once the objects arepicked up and placed in corresponding cavities or when the time foreseen for the run ends. Notethat the placement of the object in the cavity is finished when the object is fallen into the cavity (i.e.at least some part of the object has to touch ground floor underneath the cavity).

All objects to be transported in a run of a team and the corresponding cavities share the sameorientation, either horizontal or vertical. This may vary between different teams and different runs.

An example task specification is given here:

PPT<S1,(M20, S_40_40_G),S2>

where S1 is the location where to pick up the objects M20 and S_40_40_G. The objects need to beplaced then in the corresponding cavities at location S2.

6.5.5 Complexity Levels

All Complexity Options from BMT apply.

● PPT Orientation Complexity (bonus factor = 0.2):○ The cavities can be placed in all orientations.

● PPT Rotation Complexity (bonus factor = 0.2):○ The cavities can be placed in all orientations.

6.5.6 Rules


33

The following rules have to be obeyed:●The order in which the teams have to perform will be determined by a draw.●A team has a time period of 8 minutes to complete a run.●At the beginning of a team’s period, the team will get the task specification. ●A manipulation object counts as successfully grasped as specified in Section 5.5.3.●An object counts as placed correctly if it fell through the correct cavity and touches the groundbeyond. It may happen that an object blocks the cavity for the next object, e.g. by standingupright on the floor. In that case a referee may remove that object (which remains to count as asuccessful drop). If the referee is not able to do so and the robot places another object into theblocked cavity, it counts as a correct placement if it would have been successful without theblocking object.●The time is stopped when the robot has placed the last object correctly. If a team cannotcomplete the task within the specified time, the run will be stopped after it is exceeded.


● 50 points are awarded for successfully grasping an object.● 100 points are awarded for successfully placing a manipulation object into the correct

cavity.● 50 points are awarded if the task specification has been completely fulfilled. The task is

considered as fulfilled if all objects have been dropped in the right cavity. The robot doesnot have to leave the arena.

● a penalty of -50 points is given for each object which has been dropped into the wrongcavity

● The reached points of a test will be multiplied with a defined complexity factor depending onthe previously chosen complexity level.

6.6 Final (Example)


34

6.6.1 Purpose and Focus of the FinalThe purpose of the Final is to show what each robot can do in combination. The Final test will be acombination of the other tests. To show how the final test could look like the following is added:

6.6.2 Example Scenario

Basically the scenario is a BTT with a PPT and Navigation Complexity included, the following adjustments have been made:

● Time limit is 10 min. ● 5 Objects will have to be transported in total. ● Two objects will be used twice. ● 3 Service- and 3 Destination-areas will be used. One dest. area is a PPT area.● The robot will have to start at the entrance, and to complete the run exit through the exit.

6.6.3 Complexity

From BNT: ● Obstacle complexity● Barrier Tape complexity

From BMT:● Manipulation object complexity● Decoy object complexity● Orientation complexity● Rotation complexity● Position complexity

From CBT: (if used)● Speed Complexity

From PPT: ● PPT Orientation Complexity● PPT Rotation Complexity

6.6.4 Scoring

● 50 points for (the first time) reaching a service or destination area. An area counts asreached when any part of the robot is within 1m of the area and the robot is orientedtowards the area. The robot does not have to stand still.

● 50 points for grasping a (correct) object. Grasping is defined like in BTT. ● 50 points for transporting and putting to the correct destination area. ● 50 points for placing into the correct cavity of a PTT plate. ● If an object is placed into the wrong cavity or to the wrong area no penalty points apply. ● 50 points for completing the test, + time bonus points according to rules.● Other penalty points like losing objects etc. apply according to the BTT test, if not stated

otherwise above.


Documents

Rule Book - robocupatwork.orgrobocupatwork.org/download/rulebook-2015-06-04.pdf · Rule Book Version 2015.2 Juni 4, 2015 Status: China 2015 Version Authors: Walter Nowak Gerhard Kraetzschmar