recent advances on locomotion mechanisms of hybrid mobile ... · recent advances on locomotion...

Recent Advances on Locomotion Mechanisms of Hybrid MobileRobots

SHUN HOE LIMUNIVERSITI MALAYSIA SABAHEvolutionary Computing Laboratory

Faculty of Computing and InformaticsKOTA KINABALU, MALAYSIA

[email protected]

JASON TEOUNIVERSITI MALAYSIA SABAHEvolutionary Computing Laboratory

Faculty of Computing and InformaticsKOTA KINABALU, MALAYSIA

[email protected]

Abstract: Many types of hybrid mobile robots that combine the characteristics of wheeled robot and legged robothave been developed in the past two decades. This paper is aimed at presenting a survey on various hybrid mobilerobots based on their respective implemented locomotion mechanism. The survey is done on several recentlydeveloped hybrid mobile robots by inspecting the design concept of their locomotion mechanism. Besides that,this paper also discusses the factors that influence the design of a robust robotics platform which are important toconsider when designing a robot. This work will be useful as a preliminary reference point for those who want todesign a hybrid mobile robot.

Key–Words: hybrid mobile robots, robot locomotion, wheeled robots, autonomous robots

1 IntroductionSince the first implementation of mobile robots inWorld War II, mobile robotics has become an ex-tremely popular research topic. By definition, a mo-bile robot is a machine with the capability to move ina given environment. In other words, mobile robotsare able to move around in a specified environmentand are not just fixed to one physical location. Mo-bile robot is a field of great interest in robotics asit has a close interaction with environment. Mobilerobotics can be utilized in a wide range of applica-tions. For example, service industry, military deploy-ments, manufacturing, cleaning, entertainment and re-mote exploration, especially in search and rescue op-erations where human lives can be endangered.

For ground mobile robots’ locomotion, wheelsand legs are the two common adopted methodolo-gies. From a biological perspective, land animals withtheir sturdy legs are able to move over uneven terrainssmoothly and rapidly after a long evolvement process.On the other hand, during pre-historic times, humansinvented wheels that were specialized in rolling toassist on ground locomotion. The excellent perfor-mance of wheels in both power efficiency and travel-ling speed can scarcely be achieved by legged mech-anism. A hybrid platform with the combination ofleg and wheel has excellent maneuverability on flatground and uneven terrain. Therefore, a hybrid plat-form is highly recommended for general indoor andoutdoor environment operations as it is the trend for

”future” mobile platforms [1].Mobile robots have been showing a great success

in the real world implementation. For the first time,robots were assisting in an actual urban search andrescue mission of the World Trade Center tragedy on11 September 2001. The team assisted by search andrescue robots had succeeded to discover more than 10victims which are more than 2 percent of total victimsdiscovered [2]. The successful involvement of mobilerobots in real life rescue mission has garnered muchattention from researchers.

In recent years, hybrid mobile robots have beendesigned for various functionality and purposes. Forexample, hybrid mobile robots designed for stairsclimbing purposes, performing jumping behavior, in-situ reconfiguring robots posture and adapting to un-even terrain, among others. In general, hybrid mobilerobots can carry out their mission better in rough ter-rain compared to traditional wheeled or legged mobilerobots. Hybrid mobile robots utilize the advantages ofboth wheeled and legged mechanisms while compen-sating the downside of each other.

There are many successful examples of hybridmobile robots which are built and designed for widerange of operations. A group of researchers froma few universities in Japan had developed a hybridwheeled-legged platform through a retracting mech-anism inspired by the armadillo [3]. The idea of a re-tractable wheeled-legged module is that the specially-designed wheels can be transformed into a legs-like

WSEAS TRANSACTIONS on SYSTEMS Shun Hoe Lim, Jason Teo

E-ISSN: 2224-2678 11 Volume 14, 2015

mechanism. PAW proposed by McGill University, isa four legs robot with wheels equipped at the end ofeach leg [4]. PAW is the first to combine wheeledmode locomotion with dynamically stable legged lo-comotion. University Lubeck in Germany developedWheeHy which is capable of doing in-situ reconfig-uration of its posture [5]. One of the key featuresof WheeHy is that the robot can perform adaptationof its posture during its traversal over uneven ter-rain. National Taiwan University proposed a Quat-troped platform with hybrid legged-wheeled locomo-tion. The proposed system utilizes a transformationmethod where the morphology of its wheels can bedirectly transformed into legs.

The aim of this work is to primarily inform thereader of the recent developments related to hybridmobile robots from the review of numerous publishedpapers in hybrid robots, and provide readers withsome typical and promising mechanisms of hybridmobile robot research. More and more research worksare focused on hybrid mobile robots as their traversingcapabilities in various rough environments. In section2, we discuss the factors that influence the design of arobust robotics platform which are important to con-sider when designing a mobile robot platform. Vari-ous types of hybrid mobile robot mechanism are dis-cussed in section 3 and the studies on previous worksare discussed in detailed in section 4. In the last sec-tion, a conclusion is made as an overview of currenthybrid mobile robot mechanisms and suggestions areproposed for future hybrid mobile robot design anddevelopment.

2 Factors That Influence The DesignOf Robots

There are many factors that can contribute to the suc-cessful or failure of design, realization and function-ality of a robotic platform. As seen from practice, itis very difficult to design a robot that can be func-tioning in multiple scenarios and terrains with dif-ferent purposes. Commonly, robots are designed toperform specified tasks under certain environmentalconditions. Therefore, robots can have different sizesand different locomotion mechanisms depending onthe robots’ missions respectively.

2.1 Factor - SizeWhen designing robots, designers may face difficultyto decide the size of robots where bigger or smallerrobots have their own advantages and disadvantages.Bigger size robots can have more batteries, sensorsand actuators that can be put onboard. More batteries

may on the first glance mean longer run-time of therobot as it can bring along back-up power supplies.However, bigger robots have a heavier weight if com-pared to smaller sized robots. Therefore, in order tomove the heavy system, more electrical energy has tobe supplied to the actuators/motors of the system.

Besides, bigger robots have the advantages tocarry more processing elements/parts (i.e. embeddedsystems, sensors, and additional electronics) and morepayload (i.e. aid materials during urban search andrescue mission) onboard. The robustness and func-tionality of the system can be enhanced by havingmore useful processing elements or parts. Due to thelarger amount of sensors and actuators, the fault toler-ance of the robot is also increased [5]. For example, ifa sensor or actuator fails, the other sensors and actua-tors are still available, so the failure of some elementsmight not directly affect the performance of robot.

However, the bigger the system, the less agile andless maneuverable it is to go through some smallerand narrower paths. But for some rough terrains, de-pending on the locomotion mechanism implementedon the robot, the bigger they are, the easier it will befor the robot to overcome some bigger obstacles therobot, the bigger they are, the easier it will be for therobot to overcome some bigger obstacles [5]. Apartfrom this, there are robotic systems which are smallerin size. Smaller robots are preferable and suitable tocarry out missions which have to go through smallerand narrower passages. However, the smaller the sizeof the legs/wheels or other traction elements by suchrobots may hinder the robots to easily overcome big-ger obstacles and traverse in rough terrain.

Smaller robots have lighter weight as one of theiradvantages. This indicates the usage of actuators orany locomotion elements that consume less electricenergy as opposed to the bigger robots. Althoughwith the drawback of fewer batteries on board, smallerrobots may probably overcome this with lower en-ergy consuming actuators. However, fewer sensorson smaller robots may indicate lower levels of re-dundancy as well as the system might provide datacollected in lower quality or with limited resolution.This is because the quality of data will be influencedby the size and number of sensors integrated on therobot. In conjunction with these drawbacks of smallersized robots, these robots can be utilized to build robotswarms. Swam robotics is an approach to the coor-dination of multi robot systems where a large num-bers of small physical robots are grouped to performcertain tasks. The robots in this way may still fulfilthe mission provided that they will cooperate togetherand interchange acquired sensor data. The failure ofsome robots may not directly affect the outcome ofmissions.


E-ISSN: 2224-2678 12 Volume 14, 2015

Lastly, smaller robots are usually less costly thanthe bigger robot systems and therefore this is one moreargument that can be considered for building up robotswarms to carry out some mission tasks [5].

2.2 Factor - Locomotion MechanismsAnother factor when designing a robot is that appro-priate locomotion mechanisms should be chosen ac-cording to the operating environment of the robot. Lo-comotion design in mobile robots can traditionally bedivided into two methods: wheeled (tracked mecha-nisms can be included in this category) and legged.

Wheeled robots have the characteristic that theycan traverse a longer distance with a faster speed withtheir wheels than legged robots. Besides that, wheeledrobots are more powerful in terms of load/weight ratio[6]. This may be due to the current state of actuationtechnology where rotary actuation is more energy effi-cient and robust than the current state of linear and hy-draulic actuation technology when comparing the dis-tance traversed with such actuators [5].Additionally,it is easier and less complex when designing the con-troller of a wheeled robot. By having those advan-tages, it directly benefits the development cost of awheeled robot whereby it is cheaper than buildinga legged robot. However, there are limitations forwheeled robots as wheeled robots generally havingdifficulties when traversing over a rough terrain i.e.with obstacles, steps, discontinuous contact surface,among others.

On the other hand, when looking into nature, an-imals with legs are capable to perform multiple be-haviors i.e. walking, running, and jumping over dif-ferent variety of terrains. This has been the inspira-tion for researchers to develop legged robots in pur-suing the excellent locomotion behaviors. In gen-eral, legged robots provide a flexible adaptive mobil-ity in unstructured environment and a better perfor-mance while traversing over rough terrain. Althoughthere are ongoing researches on developing or study-ing legged robots, the state of the art of nowadays forlegged locomotion mechanism are still less efficientthan the natural ”way” of leg motion seen from ani-mals [5].

There are some techniques nowadays which try tomimic some locomotion processes seen in nature anddevelop more efficient locomotion systems, but it canbe still concluded that current legged robots nowadaysare less energy efficient than the wheeled robots [5].However, legged robots are still having better mobilityin rough terrain since they can use isolated footholdsthat optimize support and traction, whereas wheels re-quire a continuous path of support [7].

Since there are advantages for both wheeled and

legged locomotion mechanisms, here it raises the ideaof combining the advantages of each other into a sin-gle platform: hybrid mobile robots. In general, hy-brid mobile robots are integrated with both legs andwheels. There are few types of hybrid mechanismsthat have been developed and the mechanisms are dis-cussed in detailed in the next section, section 3.

3 Hybrid Mobile Robot LocomotionMechanisms

Fig. 1: Mobile Robots Locomotion Mechanism Cate-gorization

Fig.1 shows the overall summary of mobile robotslocomotion mechanism classification. As hybrid mo-bile robots are employing the merits of both tradi-tional wheeled and legged locomotion mechanisms,hybrid locomotion mechanism seems to be the trendin designing mobile robots. Thus, various hybrid lo-comotion mechanisms have been invented and devel-oped over the last decade. Basically, hybrid wheel-leglocomotion mechanisms can be categorized into threecategories as shown in Fig.2. From the figure: (a)legged mechanism attached with a wheel at its footend or known as articulated-wheeled, (b) independentwheel and leg modules on the body of the mobilerobot, and (c) reconfigurable or transformable wheelmodules which can be transformed into leg modulesand vice versa.

3.1 Leg with Wheel at the End (Articulated-Wheel) Hybrid Locomotion Mechanism

With wheels attached at the end of each legs, thiskind of hybrid mechanism helps a traditional leg robot


E-ISSN: 2224-2678 13 Volume 14, 2015

Fig. 2: Categories of Hybrid Wheel-Leg Robot (Usedwith permission from [3])

to overcome their complex and slow walking mecha-nism. On the other hand, the articulated wheels allowthe hybrid mobile robot to traverse through unevenand also discontinuous contact point terrain where atraditional wheeled robot might have difficulty to gothrough. Besides that, there are researchers who havebeen making use of this mechanism to build robotswith bounding gaits.

Basically, a hybrid mobile robot with this loco-motion mechanism may have two switchable modesof locomotion which are wheeled mode and leggedmode. However, different sizes of wheels give meritand demerit to each mode [3]. When it is on wheeledmode, bigger wheels are preferable than small wheelsas bigger wheels allow the hybrid mobile robot toclimb or traverse over a single step. However inlegged mode, smaller wheels are better than bigwheels. Due to its small footprint, the hybrid robotcan choose a good contact spot of the foot on uneventerrain with gaps [3].

There are a lot of hybrid mobile robots that havebeen built with this kind of hybrid locomotion mech-anism. For example, Roller Walker [8] which has legswith passive wheel at the end enables it to switch be-tween leg locomotion and a roller skating motion. Anovel leg-wheel hybrid stair-climbing vehicle ”ZeroCarrier” [9], which consists of eight unified prismatic-joint legs, four of which attached with active wheelsand other four attached with passive casters. Hylos[10] utilizes active suspension-leg mechanism withits four wheels providing it the ability to reconfig-ure its posture when traversing through rough terrain.A bounding gait robot PAW [4] also employs activewheels at the distal end of each leg.

3.2 Independent Leg and Wheel ModulesHybrid Locomotion Mechanism

This hybrid locomotion mechanism can be classifiedas wheeled mechanism with the assistance of leg mod-ules or in the opposite way which is a legged mech-anism with the assistance of wheel modules. Basi-cally, for wheeled mechanism with the assistance ofleg modules hybrid locomotion mechanism, the robottraverses with wheel modules while making use of leg

modules to increase the maneuverability of the robotin rough terrain. With leg modules embedded, a tradi-tional wheeled robot may have the ability to climb orpass over obstacles.

For legged mechanism with the assistance ofwheel modules hybrid locomotion mechanism, therobot moves by its leg modules with the support ofpassive wheels for stability purposes. Static and dy-namic stability is provided by the set of wheels, whilelocomotion is still mainly dependent on the legs’ mo-tion. By replacing some legs of an ordinary leggedrobot into passive wheels, the new hybrid mobilerobot can exploit the advantages of both legged andwheeled robot. With the passive wheels, the robotcan move faster and carry more weight with a sim-pler controller [6]. At the same time, the motion ofthe legs provides the ability for climbing/overcomingobstacles and better maneuverability in rough terrain.

For example, Chariot III [11], a leg-wheel robotwith four legs and two independent wheels. Thereare two operation modes for Chariot III which arewheeled mode and leg-wheeled mode. The hybridlocomotion of Chariot III is designed for moving onunexplored rough terrains. Wheeleg [12] is a wheel-legged robot which has two individual rear wheelsand two front legs with three degrees-of-freedom.A bio-inspired hybrid leg-wheel robot built by KingMongkut’s Institute of Technology North Bangkok[13] also implemented two front legs and two rearwheels design. The biological principle of swingingtwo front legs alternatively during the walking of in-sects was the inspiration for the robot’s design. Therobot movement is propelled by the two front legspushing the wheels to go forward and backward.

3.3 Reconfigurable or Transforamble WheelHybrid Locomotion Mechanism

Distinct from both previous hybrid locomotion mech-anisms which have separate wheels and legs mecha-nisms, this mechanism of hybrid locomotion utilizestransformable or reconfigurable wheels. The trans-formable wheels can change into legs when leg loco-motion is preferable for example traversing in roughterrain.

For example, a hybrid legged-wheeled platformQuattroped was introduced by National Taiwan Uni-versity. The robot implements a transformationmethod where the wheels of the robot can be di-rectly transformed into 2 degree-of-freedom legs. Anarmadillo-inspired wheel-leg robot was proposed byOsaka University. Each wheel of the robot was builtwith joints and can be turned into a leg by bending thejoints in a reversed direction [3].


E-ISSN: 2224-2678 14 Volume 14, 2015

4 Study On Recent Developed Hy-brid Wheel-Leg Mobile Robots

The idea of hybrid wheel-leg locomotion has beenproposed several decades ago when researchers weretrying to compensate the pros and cons of wheeled andlegged locomotion. Thus, there are numerous hybridwheel-leg mobile robots that have been developed andbuilt up to today. However, in this section, we are pre-senting the discussion on the latest hybrid wheel-legmobile robots which had been developed in last fewyears. The discussion will be categorized accordingto the hybrid locomotion mechanism classification.

4.1 Leg with Wheel at the End (Articulated-Wheel) Hybrid Locomotion Mechanism

4.1.1 Bounding Gait in a Hybrid Wheel-LegRobot

The PAW (Platform for Ambulating Wheels) robot,an articulated suspension system, implements the legwith a wheel at the end hybrid locomotion mecha-nism. PAW was developed by Centre of IntelligentMachines from McGill University with the supportof Autonomous Intelligent Systems Section DefenseR&D Canada [4] [14].

Fig. 3: The PAW robot with bounding gait (Used withpermission from [14])

PAW as shown in Fig.3, is a four-legged robotwhere each of its leg is attached with wheel. PAWuses SCOUT II’s frame but with a terser and lighterversion. However, each leg of PAW is equipped withan active wheel instead of a passive wheel. The pri-mary operation mode of PAW is wheeled mode. Inthis mode, all the wheels of the robot can be reposi-tioned by the four hip motors. While in legged mode,each wheel is actively locked, allowing PAW to per-form dynamic behaviors such as bounding and jump-ing. By equipping legs with repositionable wheels,

PAW is able to perform more advantageous turning,braking and bounding movements.

PAW implements an altered edition of the stan-dard differential steering method to drive the turn-ing of the robot. The traditional differential steeringmethod is driven by changing the speed of one sideof the wheel of the robot while the position of legs isfixed whereas PAW changes the position of its wheelsby moving its legs to lower shear forces on them. Bybringing the inner legs of the turn closer while keep-ing both of the outer legs upright, the centre of massof the robot will be lower and thus lean the robot intothe turn.

Besides turning, PAW is able to brake or performstopping without pitching over. Braking in a suddenor inappropriate legs angle can result in pitching mo-tion. The legs of PAW are places in a sprawled postureduring forward and reverse driving, which is aboutplus or minus 11.5 degrees with regard to the robotbody’s vertical reference. During braking, the kineticenergy of the robot can be dissipated by motors usinglow gain PID controllers.

For the bounding gait of PAW, two separate statemachines are used to drive the gait where one for thepair of rear leg and the other one for the front. Thebounding behavior of PAW starts from a standing po-sition and continues with a merging of open loop leanback and kicking movements. There are three phasesfor a single bounding behavior, which are the flightphase, stance retraction phase and stance break phasewhich can be seen in Fig.4.

Fig. 4: Phases in rear and front legs’ bounding statemachines

During the flight phase, a position-based PID con-troller is used to control the legs of the robot in orderto obtain the desired touchdown angle. A constant de-sired stance torque is used during the stance retrac-tion phase, in order to reposition the robot into takeoffangle. After reaching the takeoff angle, stance brakephase will take place. The bounding state machine


E-ISSN: 2224-2678 15 Volume 14, 2015

takes care of the switching procedure between thesemodes.

4.1.2 Reconfiguration and Obstacle NegotiationMethods on Hybrid Leg-Wheeled Robot

Graduate University of Chinese Academy of Sciencesin cooperation with State Key Laboratory of RoboticShenyang proposed a reconfiguration control methodfor a hybrid leg-wheeled robot [16]. They used a sixleg-wheeled robot as their platform. 12 individual mo-tors are used to drive the six legs and six wheels inde-pendently. Passive suspensions are installed at eachof the legs, which will later be used as an informationcollector of contact state between wheels and terrain.Besides reconfiguration method, they also exerted theobstacle negotiation ability to the leg-wheeled robot.

Their proposed reconfiguration method consistsof three stages. Firstly, according to the power ex-erted on legs, the touching status of the wheel on theterrain can be estimated. After that, information aboutthe current configuration of the robot is gathered inorder to compute the expected leg angles. Lastly, thelegs are adjusted according to the angles calculated insecond stage to obtain the expected configuration.

The most important feature that differentiatesa hybrid legged-wheeled robot from a traditionalwheeled robot is its ability to negotiate obstacles [16].The proposed obstacle negotiation ability can mainlyhandle standard obstacles, such as ridges and ditches.By using laser range finder with obstacle detecting al-gorithm, the robot can determine its ability to strideover the approaching obstacle. The robot will thenregulate the heading to locate an appropriate strid-ing direction if the obstacle can be stridden. Oth-erwise, the robot will activate an obstacle avoidancestrategy to bypass the obstacle. As crossing ditchesand ridges requires an association of moving upwardand downward procedures, two control strategies havebeen proposed in order to enable it to climb with anupward step and a downward step.

Climbing an Upward-Step

The robot will measure the gradient of the slope when-ever the robot approaches an upward-step. The robotwill climb on it directly by implementing the recon-figuration method if the robot can pass through thestep (small gradient). Otherwise, the robot will usethe control strategy as in Fig.5 to climb over it.

Firstly, the robot will reconfigure itself accordingto the height of the upward step as shown in Fig. 5.1.With the support of an odometer, the robot will moveforward in order to put the front wheel on top of thestep. After the front wheel is put on top of the step,

Fig. 5: Reconfiguration methods for climbing the up-ward step

the angles of the front legs and rear legs are adjustedas in Fig. 5.2. The robot will move forward againto place the middle wheel on top of the step. Afterthat, the rear leg will be pulled onto the step then onlythe robot will be recovered back to the original robotconfiguration.

Climbing an Downward-Step

Similar with the climbing upward step strategy, therobot will measure the gradient of the slope whena downward-step is approached. The robot will godown the step directly if the gradient is small enough.Otherwise, the control strategy as in Figure 6 will beimplemented to climb down from the step.

Fig. 6: Reconfiguration methods for climbing thedownward step

Firstly, the robot will be reconfigured as in Fig.6.1 and move forward. The robot will move forwarduntil the front wheel goes off from the step. In or-der to identify whether the front wheel is out from thestep, the velocity of the moving robot and the revolu-tion rate of the front wheel will be inspected whilethe robot is moving forward. This can be inferredwhere the moving speed of the robot is much smallerthan the revolution rate of the front wheel. When thefront wheel is found going off from the step, the robotwill move forward with the assistance of the odometer.


E-ISSN: 2224-2678 16 Volume 14, 2015

The robot will then stop after marching a certain dis-tance. After that, the angle of the front leg is adjusted.With the similar method that has been descried previ-ously, the robot will advance forward along with theinspection on whether the middle wheel is going offfrom the step. When the middle wheel is found goingoff from the step, the robot is reconfigured as in Fig.6.3. Finally, the robot will move forward while puttingdown the rear wheel onto the floor and the robot is re-covered to the original marching configuration.

4.1.3 In-situ Reconfigurable Hybrid Wheeled-Legged robot - WheeHy

The Institute of Computer Engineering from Uni-versity Lubeck Germany proposed the design ofa wheeled-legged hybrid robot platform namedWheeHy [5]. WheeHy robot is entitled of adapting itsposture to uneven terrain which it is traversing overand performing in-situ reconfiguration of its posture.

The design of WheeHy is different from commonhybrid wheeled-legged robot where it is a three leggedwheel-legged robot instead of quadruped robot whichcan be seen in Fig.7. The purpose of having a threelegged robot design is to lower the weight of the robotin comparison to the quadruped robot, and in the sametime the stability and maneuverability are not drasti-cally decreased.

Fig. 7: Front view and side view of WheeHy (Usedwith permission from [5])

WheeHy was designed with ”star” like wheels inorder to lower its weight and at the same time to haveenough supporting elements to cope with the weightof the robot and dynamics introduced when traversingover different terrains. Rubber elements are integratedat the end of each part of the ”star” like wheels for abetter grip on different kinds of surfaces. However,there are wheel ”adapters” on each of the robot legs,this enable the wheels of the robot to be replaced withsome other type of wheels if required.

The robot’s control architecture provides severalinterfaces which including the control for the legs andwheels; balancing of the robot over uneven terrain;check for the need and perform reconfiguration anddata logging. There is a network interface provided,thus WheeHy can be remotely controlled. Fig.8 showsthe control architecture of WheeHy.

Fig. 8: Control architecture of robot WheeHy (Usedwith permission from [5])

For a balancing strategy over an uneven terrain,inclination sensors are used to obtain the interactioninformation of wheels and surface. Depending on thevalues read from the inclination sensors, the robot legswill change in upward or downward direction in or-der to adapt to the terrain conditions and maintain aparallel position of the robot body with respect to theterrain. This balancing capability of the robot allowscrossing over an uneven terrain with more stabilityand therefore lowering the risk of tipping over in un-even terrain.

Although the balancing capability of the robot al-lows it to mitigate potential tip over situations, therobot is still enhanced with a new feature which isthe reconfiguration capability. When the robot is tip-ping over, it can realize the tilt position with the pro-vided onboard accelerometer sensors that it has tum-bled over. The robot can recognize that it has tippedover by reading the current values and comparingthem with the ”normal” body position accelerometervalues which are known in advance. Then the recon-figuration strategy is activated by stopping the move-ment of the robot wheels at first. After that, the legsof the robot are stretched out in parallel to its bodyposition and moved down on the appropriate side ofthe body. In that manner, WheeHy is able to stand upagain and continue with its mission.


E-ISSN: 2224-2678 17 Volume 14, 2015

4.2 Independent Leg and Wheel ModulesHybrid Locomotion Mechanism

4.2.1 A Bio-Inspired Hybrid Legged-WheeledMobile Robot

A bio-inspired hybrid legged-wheeled mobile robothas been developed by the Mechanical EngineeringDepartment of King Mongkut’s Institute of Technol-ogy North Bangkok [13]. They used the biologicallyinspired kinematics of insect’s legs as the leg designof the hybrid leg-wheel robot. They utilized one ofthe key features from the construction of insect legwhich is the multi-segmented nature where segmentjoints consist of single or multiple degrees of freedom.Thus, the robot in their work implemented the biolog-ical principles of swinging the two front legs alterna-tively during the walking of insect as the inspirationfor the robot design.

The robot has two front legs that are functioningsimilarly with insect legs. However, the robot legs arekinematically simpler than the insect in order to sim-plify the mechanical design. Both legs of the robotwere designed with two degrees of freedom. With thisleg design, the robot will have the capability to navi-gate over rough terrains and move over large obstacleswith a faster speed and less energy. There are two pas-sive wheels attached at the back of the robot. Thus, therobot movement is propelled by the two front legs thatpushed the wheels to go forward and backward. Fig.9shows the model of the robot.

Fig. 9: Model of Bio-Insipred Hybrid Leg-WheelRobot (Used with permission from [13])

Fig.10 and Fig.11 show the movement of the twofront legs of the robot. The leg can move in a total of90 degreees from the top view angle as well as fromthe front view angle.

Fig. 10: Front View of Leg Movement of Bio-InspiredHybrid Leg-Wheel Robot (Used with permission from[13])

Fig. 11: Top View of Leg Movement of Bio-InsipredHybrid Leg-Wheel Robot (Used with permission from[13])

The controller developed for the hybrid leg-wheelrobot front legs was based on common feed-forwarddesign. The controller is programmed on a basicStampbox microcontroller by Parallax. There are to-tal four motors are embedded in the robot where twomotors on each leg. Since the wheels movement is ranby the driving force of the two front legs, no motor isrequired for wheels. The Stampbox will take the feed-back signals from the four joint sensors which are po-tentiometers and also from foot sensors to determinethe foot state whether is on or off the ground. Thenthe Stampbox will send either a clockwise rotation orcounterclockwise rotation signal to motors.

4.2.2 A Two Legs and Two Independent WheelsHybrid Robot, Wheeleg

In the DEES Robotic Laboratory of University ofCatania, the practical design of the robot Wheeleg wasrealized and built [12]. Wheeleg is a robot with twoindividual rear wheels and two front legs. The twocylindrical type front legs of the robot are pneumati-cally actuated with three degrees-of-freedom. Whilethe two back wheels of the robot are individuallydriven by two motors. Fig.12 shows the hybrid robot,Wheeleg.

With the design of the rear wheels, mostly allof the robot weight is supported by the rear wheels.While the front legs are designed for improving grippurposes which also enable Wheeleg to overcome ob-stacles. The design of Wheeleg robot allows it to havean improved maneuverability on rough terrain, if com-


E-ISSN: 2224-2678 18 Volume 14, 2015

Fig. 12: The hybrid robot, Wheeleg (Used with per-mission from [12])

paring to an ordinary wheeled mobile robot. At thesame time the movement of Wheeleg is also faster,more stable and easier to control than an ordinary fourlegs mobile robot.

The design of Wheeleg acts as a beneficial so-lution for operation in a simpler environment wherelegged mobile robots are not required, but at the sametime a better surface gripping is preferable. However,according to the developer, there are drawbacks fromthe design of Wheeleg. For example, insufficient pres-sure on legs causing traction difficulty, the constrainton static and dynamic stability when traversing overrough terrain, and a more complicated control systemis needed in order to drive the wheels and legs simul-taneously.

Wheeleg robot has a total of eight microcon-trollers, six for controlling the pistons and the othertwo for the set of wheels. The overall control super-vising and user interfacing is done by a microproces-sor. There are four digital valves in totals which arejoined with each pneumatic cylinder of both of thelegs where two valves are for air inlet and the othertwo are for outlet. Pulse width modulation (PWM)signal is used to control the digital valves which canbe generated by controllers. There is a touch sensormounted on each foot in order to determine which footis on the surface of terrain. A linear potentiometer ismounted on each joint of the leg in order to give feed-back signal of the leg position to the pneumatic con-troller.

While for the wheels, each wheel is actuated bya standard brush DC motors with gear reducers. Lowcost standard brush DC motors were purposely chosenin order to decrease the system cost. Each motor iscontrolled by a different controller respectively whichis coded with the position feedback encoder. Besides

that, the controller is also connected to the master pro-cessor for exchanging commands purposes.

4.3 Reconfigurable or Transformable WheelHybrid Locomotion Mechanism

4.3.1 A Leg-Wheel Hybrid Mobile Platform withTransformable Wheel Morphologies

A four legged legged-wheeled hybrid platform wasproposed by the Department of Mechanical Engineer-ing from National Taiwan University [1]. Distinctfrom the other leg-wheel hybrid mobile robots whichare mostly having separated mechanism for wheelsand legs or articulated wheels, this robot implementsa transformation mechanism where each wheel of therobot can be directly transformed into two degree offreedom legs. The transformation mechanism is actu-ally changing the wheels which are in around shape,into legs by breaking up the round wheel into two halfcircles and combining it as a leg. The robot in bothlegged and wheeled can be seen in Fig.13 below.

Fig. 13: The Leg-Wheel Hybrid Robot in Legged andWheeled Mode (Used with permission from [1])

The crucial part in the design of this robot is thetransformation mechanism which enables the robot todeform a particular part of the robot morphology tofunction as legs or wheels. A wheel normally con-sists of a rotary axis and a spherical rim where therotary axis is situated at the middle of the sphericalrim. A ”hip” joint is the point where the rotary axislinks with the mobile platform. In wheeled locomo-tion, the point of the hip is fixed where the distancefrom the hip to the touching point with ground is theradius of the circular rim. However in legged locomo-tion, the connection between the hip and the touchingpoint with ground is not confined. Therefore, the loco-motion can be switched from wheeled mode to leggedmode by moving the hip point out from the middle ofthe rim.


E-ISSN: 2224-2678 19 Volume 14, 2015

As we can see in Fig.13, the robot was built upwith four wheels. In wheeled mode, the locomotionmechanism of the robot is similar to a 4-wheel-drivevehicle. The robot can be moved forward or backwardwhen rotation motions are activated at the hip joints.The turning motion of the robot is achieved by steer-ing the front wheels according to Ackermann steeringgeometry.

Fig. 14: leg mode locomotion (a) walking on roughterrain; (b) climbing across obstacles; (c) climbingstair ascent (Used with permission from[1])

In legged mode, the hip point is shifted closer tothe rim after the rim of the robot is folded in half. Therobot is then turned into a four-legged robot as illus-trated in Fig.13. In legged mode, the robot is capableto traverse through rough terrain more smoothly thanwheeled mode. Besides that, the robot is able to climbacross large obstacles and also ascent or descent stairs.The leg mode locomotion on various terrains is shownin Fig.14.

4.3.2 Armadillo-Inspired Wheel-Leg RetractableRobot

The Department of Mechanical Engineering fromGraduate School of Engineering, Osaka Universityproposed an improved hybrid wheeled-legged plat-form through a retracting structure inspired by the ar-madillo [3]. The robot is a Quattroped with four re-tractable legs which can be transformed into wheels.The proposed retractable mechanism comprises alarge wheel diameter to realize a high ability on climb-ing obstacles.

The idea of the retractable wheeled-legged mod-ule is illustrated in Fig.15. The wheel contains all thejoints of a leg; hence, it is able to achieve a larger di-ameter for climbing obstacles.

Fig. 15: Retractable Wheel-Leg Module (Used withpermission from [3])

The proposed retractable wheeled-legged moduleenables the robot to have a better maneuverability. Itis easier for the robot to climb small single-step andobstacles due to its large diameter of wheel when therobot is in wheeled mode. On the other hand, whenin legged mode, the robot has the ability to choose theposition of the foot end on uneven terrain by fully uti-lizing the small foot print of its leg. Fig.16 shows theprototype of the retractable wheel-leg hybrid robot.

Fig. 16: The Prototype of the Retractable Wheel-LegHybrid Robot (Used with permission from [3])


E-ISSN: 2224-2678 20 Volume 14, 2015

The retractable wheel-leg module enables therobot to have another application which is the rollinggrasping motion. This module can act as gripper withsome joints when it is on legged mode to performgrasping operation as shown in Fig.17(a). While onwheeled mode, this module can act as gripper with alarge roller as illustrated in Fig.17(b).

Fig. 17: Grasping Motion with Joints and Rolling(Used with permission from [3])

4.4 SummaryThere are total of seven hybrid mobile robots thathave been reviewed in the last section. The reviewis focused on the design of their locomotion mech-anisms and is categorized in three categories, whichare leg with wheel at the end (articulated-wheel), inde-pendent leg and wheel modules and reconfigurable ortransformable wheel mechanisms. A summary of thestructure and features of the reviewed robots is listedin table 1 (Appendix).

5 CONCLUSION AND WAY FOR-WARD

This paper classifies mobile robots into three cate-gories which are wheeled robots, legged robots andhybrid mobile robots. For hybrid mobile robots, thereare three common types of locomotion mechanism:legs with wheels at the end (articulated wheel), inde-pendent leg and wheel modules, and reconfigurable ortransformable wheel modules. The main part of thispaper is to present a survey on recently developed hy-brid mobile robots by inspecting their design conceptsand control methodology. Apart from that, this paperalso presents a discussion on the factors that influencethe design of a robust robotic platform which are theimportant criteria in designing a hybrid mobile robot.

As has been reviewed, numerous hybrid mobilerobots have been proposed and developed. How-ever, as far as we are aware, most of the hybrid mo-

bile robots are manually designed where the designersmust have the preliminary knowledge of the interac-tion between the robots with the environment. Theuse of artificial evolution for the automatic generationand synthesis of controllers and/or morphologies forrobots is one of the more recent methods in devel-oping robots [17][18][19]. By implementing evolu-tionary algorithms in designing a robot, an optimizedcontroller and/or morphology can be obtained whereat times, these evolved solutions might be beyond thedesigners’ design capability.

A co-evolution approach to sensor placement andcontrol design for robot obstacle avoidance has beenproposed by Wang and others [20]. Obstacle avoid-ance can be considered as one of the most importantfeatures of an autonomous mobile robot. Previously,numerous obstacle avoidance approaches were basedon a specific robot hardware design and subsequentlyexperimented with to obtain the optimal controller de-sign. The sensor placement for the robot is based onthe designers’ experience or common sense which ishard to determine as optimal. By looking into naturalsystem, we can find animals that co-evolved their sen-sor systems (physical sensory attributes) together withtheir control (neural) systems when they were tryingto adapt to the environment. Thus, a co-evolutionaryapproach would appear to be highly beneficial as wellin the case of designing hybrid robots. The selec-tion and placement of sensors in addition to the de-sign of a suitably integrated control system for hybridrobot morphologies, which arguably are more com-plex and complicated than conventional wheeled orlegged robots, could be co-evolved in this case.

Similarly, simulated robots (creatures) had beensuccessfully evolved by evolving both of their mor-phologies and controllers [21]. The evolved robotshave the capability to traverse on flat and rough ter-rains. The robots are evolved through a developmentalprocess which takes place in time and space. Duringthe process, the robots are achieved through a progres-sive addition of both regulatory substances and struc-tural parts. The robots were built up with distributedcontrol systems. With a few independent neural con-trollers that are embedded in different parts of therobot which can only access the local sensory infor-mation, these will respectively form the overall con-trol system. Analysis showed that the performancesof the evolved robots were improved with respect totheir capability to move on a flat terrain by increasingthe complexity of the environment in which the robotswere being simulated in. Again, such an approach todesigning hybrid mobile robots could be benefit fromsuch an evolutionary methodology. Independent neu-ral controllers could be evolved or co-evolved to func-tion within each individual articulated portion of the


E-ISSN: 2224-2678 21 Volume 14, 2015

hybrid mobile robot’s morphology.However, there is a critical issue in evolution-

ary robotics where very often robots that are evolvedin simulations are inefficient when transferring to thereal world. This transfer problem is called reality gap[22] which is the main cause that are hindering theuse of evolutionary robotics for practical robotic ap-plications. Koos and others recently highlighted thatthere is a conflict between the efficiency of the solu-tions in simulation and their transferability from sim-ulation to reality [23]. They hypothesized that the so-lutions with best efficiency in simulation usually uti-lize badly modeled phenomena in achieving high fit-ness scores. They proposed transferability approach,where a multi-objective formulation of evolutionaryrobotics is utilized where two main objectives are op-timized via a Pareto-based multi-objective evolution-ary algorithm. The two main objectives are the fit-ness of solutions evolved in simulation and the trans-ferability of the solutions to the real world. Theyhave also suggested a simulation-to-reality (STR) dis-parity measure method to estimate the transferabil-ity objective. With the transferability approach, theyhave succeeded in finding efficient and good transfer-able controller within a very short duration of 10 ex-perimental runs after transference onto the physicalrobot. Therefore, a multi-objective approach againcould be considered in the case of evolving hybridmobile robots in order to surmount this transferenceproblem since an evolutionary approach to the designof such robots would require extensive simulation dur-ing the evolutionary optimization runs. Additionally,a multi-objective approach could be used not only toovercome the simulation-reality gap as a bi-objectiveproblem but could also be extended to three and moreobjectives to include additional design criteria suchas complexity, energy efficiency, and heterogeneity ofmorphologies.

Preliminary result of our first experiment in op-timizing the morphology of a six legged-wheeled hy-brid mobile robot shows that evolutionary algorithmcan be implemented in designing robots [24]. In theexperiment, the morphology of a six legged-wheeledhybrid mobile robot is evolved with single-objectiveevolutionary algorithm. The evolving parameters arethe radius of wheels, length of legs, and size of bodywhich are to be optimized in the evolution in orderto produce a smaller robot with the ability to performobstacle climbing motion. After the evolution simu-lation, the fittest robot is transferred into real worldwith 3D printing fabrication. Fig.18 shows the fittestrobot in simulation and Fig.19 shows the fabricatedrobot. Further investigation on the evolution withmulti-objective evolutionary algoirhtm and evolutioninvolving more parameters will be carried out.

Fig. 18: Fittest Robot Obtained from the EvolutionSimulation

Fig. 19: Fabricated Robot with 3D Printing Technol-ogy

In conclusion, evolutionary robotics have beenshowing a great success and getting more and more at-tention among developers in robotics field. The possi-bility to develop undiscovered potential of evolution-ary robotics is ultimately high and therefore more ef-fort needs to be contributed on this field for roboticsrevolution.

Acknowledgements: This work was funded underScienceFund project SCF0085-ICT-2012 granted bythe Ministry of Science, Technology and Innovation,Malaysia


E-ISSN: 2224-2678 22 Volume 14, 2015

Table 1: Summary of the structure and features of reviewed robotsCategory Robot(Developer) Structure FeaturesLeg with wheelat theend/articulatedwheelmechanism

PAW, Platform for Ambu-lating Wheels (Centre ofIntelligent Machine fromMcGill University) [4][14]

- A quadrupedal.- Each leg houses a pair of exten-sion strings and attached with ac-tuated hard rubber wheel.

- Two operating modes whichare wheeled (primary) and leggedmode.- Capable of turning, braking andbounding movements.

Six leg-wheeled robot(Graduate University ofChinese Academy ofSciences) [16]

- Six leg-wheeled robot- 12 individual motors for drivingsix legs and six wheels- Passive suspensions are installedat each leg for collecting informa-tion of the contact state betweenwheels and terrain.

- In-situ reconfiguration ability toadapt uneven terrain.- Obstacle negotiation abilitywhether to bypass obstacle orclimb over obstacle.

WheeHy (Institute ofComputer Engineeringfrom University Lubeckgermany) [5]

- Three legged wheel-leggedrobot.- Designed with star shapedwheels which can be replacedwith other types of wheel if re-quired.

- Able to maintain a parallel posi-tion of the robot body while cross-ing uneven terrain with balancingstrategy.- In-situ reconfiguration abilityenables it to reconfigure itself af-ter tipping over.

Independentleg and wheelmechanism

Bio-inspired legged-wheeled robot (Me-chanical EngineeringDepartment of KingMongkut’s University)[13]

- Designed with two front legswith two degrees of freedom and2 passive rear wheels.- The design of front legs is in-spired by the kinematics of in-sect’s legs.

- Robot movement propelled byswinging the two front legs thatpushed the passive wheels to goforward and backward.- Able to navigate over roughterrain with large obstacles withfaster speed and less energy.

Wheeleg (DEES RoboticLaboratory of Universityof Catania) [12]

- Two individual rear wheelsdriven by two motors and twopneumatically actuated front legswith three degrees of freedom.- Two digital valves are joinedwith pneumatic cylinder of eachleg where one for air inlet and onefor air outlet.

- The design of Wheeleg is abeneficial solution for operationin a simpler environment wherelegged robots are not required butat the same time a better surfacegripping is preferable.- A more complicated control sys-tem is required in order to drivethe wheels and legs simultane-ously.

Reconfigurable/transformablewheelmechanism

Transformable leg-wheelrobot (Department ofMechanical Engineeringfrom National TaiwanUniversity) [1]

- Four legged legged-wheeled hy-brid platform.- Transformable wheels which canbe transformed in legs and viceversa.- Wheels in a round shape arechanged into legs by breaking upthe round wheel into two half cir-cles and combining it.

- In wheeled mode, the locomo-tion of the robot is similar as afour wheeled drive vehicle.- In legged mode, the robot isturned into a four legged robot andcapable to traverse through roughterrain and climb obstacles andstairs.

Retractable wheeled-legged robot (Departmentof Mechanical Engineer-ing from Graduate Schoolof Engineering, OsakaUniversity) [3]

- Quattroped with four retractablelegs which can be transformedinto wheels.- Wheel with larger diameter as itcontains all of the links of a legand by bending the direction ofthe joint in reverse direction cantransform it into a leg.

- In wheeled mode, it is easierto overcome small single-step andobstacles with its larger wheels.- In legged mode, the small foot-print of the robot leg enables it tochoose the position of the foot endon uneven terrain.- Additional application which isthe rolling grasping motion withits retraction module.


E-ISSN: 2224-2678 23 Volume 14, 2015

References:

[1] S. Y. Shen, C. H Li, C. C. Cheng, J. C. Lu, S. F.Wang, P. C. Lin, Design of a Leg-Wheel HybridMobile Platform, IEEE International Confer-ence on Intelligent Robots and Systems(2009).

[2] D. Angela, Urban Search and Rescue Robots:From Tragedy to Technology, IEEE,(2002).

[3] K. Tadakuma, R. Tadakuma, A. Maruyama, E.Rohmer, K. Nagatani, K. Yoshida, A. Ming,S. Makoto, M. Higashimori, M. Kaneko,Armadillo-Inspired Wheel-Leg RetractableModule, IEEE International Conference onRobotics and Biomimetics,(2009).

[4] A. S. James, S. Inna, T. Michael, PAW: a HybridWheeled-Leg Robot, IEEE International Confer-ence on Robotics and Automation,(2006).

[5] J. Bojan, H. Martin, K. Michael, M. Erik, Designof a hybrid wheeled-legged robot - WheeHy,IEEE,(2010).

[6] C. H. Ong, H.M. Shamsudin, Amin, A Bio-logically Inspired Hybrid Three Legged MobileRobot, IEEE,(2002).

[7] C. Gianni, O. Erika, Design and Simulationof a New Hybrid Mobile Robot for Overpass-ing Obstacle, Springer Science+Business MediaB.V.,(2009).

[8] G. Endo, S. Hirose,Study on Roller-Walker,IEEE International Conference on Robotics andAutomation, vol. 1, 2032-237, (1999).

[9] J. Yuan, S. Hirose, Research on leg-wheel hy-brid stairclimbing robot,zero carrier, IEEE Inter-national Conference on Robotics and Biomimet-ics, vol. 1, 654-659, (2004).

[10] C. Grand, F. Benamar, F. Plumet, P. Bidaud,Stability and traction optimization of a re-configurable wheel-legged robot, The Interna-tional Journal of Robotics Research, 1041-1058,(2004).

[11] S. Nakajima, E. Nakano, T. Takahashi, Motioncontrol technique for practical use of a leg-wheelrobot on unknown outdoor rough, IEEE/RSJ In-ternational Conference on Intelligent Robots andSystems, vol. 1, 1353-1358, (2004).

[12] M. Lacagnina, G. Muscato, R. Sinatra, Kine-matics, dynamics and control of a hybrid robotwheeleg, Robotics and Autonomous Systems,161-180, (2003).

[13] S. Kawee, L. Sathaporn, A Bio-Inspired HybridLeg-Wheel Robot, IEEE, (2004).

[14] A. S. James, S. Inna, T. Michael, BoundingGait in a Hybrid Wheel-Leg Robot, EEE Inter-national Conference on Intelligenct Robots andSystems, (2006).

[15] I. Poulakakis, J. A. Smith, M. Buehler, Model-ing and Experiments of Untethered QuadrupedalRunning With a Bounding Gait: The Scout IIRobot, International Journal of Robotics Re-search, 24, 239-256, (2005).

[16] J. L. Lu, C. G. Bu, Study on the Mobile RobotReconfiguration Control Methods, Proceedingsof the IEEE International Conference on Au-tomation and Logistics, (2009).

[17] L. Wang, K. C. Tan, C. M. Chew, EvolutionaryRobotics: From Algorithms to Implementations,World Scientific Publishing, (2006).

[18] S. Nolfi, D. Floreano, Evolutionary Robotics,MIT Press, (2000).

[19] H. Iba, Frontiers in Evolutionary Robotics, I-Tech Education and Publishing, (2008).

[20] W. C. Wang, S. X. Yang, W. R. Shi, M. Q.H. Meng, A Co-Evolution Approach to SensorPlacement and Controller Design for Robot Ob-stacle Avoidance, Proceedings of 2004 Interna-tional Conference on Information Acquisition,(2004).

[21] M. Mazzapioda, A. Cangelosi, S. Nolfi, Evolv-ing Morphology and Controller: A DistributedApproach, IEEE Congress on EvolutionaryComputation, (2009).

[22] N. Jakobi, P. Husbands, I. Harvey, Noise andthe Reality Gap: The use of Simulation in Evo-lutionary Robotics, in Proc. ECAL, 704-720,(1995).

[23] S. Koos, J.B. Mouret, S. Doncieux, The Trans-ferability Approach: Crossing the Reality Gapin Evolutionary Robotics, IEEE Transactionson Evolutionary Computation, VOL 17, No 1,(2013).

[24] S.H. Lim, J. Teo, An Evolution Approach toOptimize the Morphology of a Six Legged-Wheeled Hybrid Mobile Robot, InternationalConference on Computational Science andTechnology, (2014).

[25] K.O. Chin, J. Teo, Artificial Neural Con-troller Synthesis in Autonomous Mobile Cogni-tion, International Journal of Computer Science,36(4):240-252 (2009).

[26] K.O. Chin, J. Teo, Evolution of RF-SignalCognition for Wheeled Mobile Robots usingPareto Multi-objective Optimization, Interna-tional Journal of Hybrid Information Technol-ogy, 2(1):43-56, (2009).

[27] A. M. Zak, O. A. Mahgoub, A. M. El-Shafei, A.M. Soliman, Design and implementation of effi-cient intelligent robotic gripper, WSEAS Trans-actions on Systems, 9(11), 1130-1142, (2010).


E-ISSN: 2224-2678 24 Volume 14, 2015

[28] G. Tont, L. Vldreanu, M. S. Munteanu, D. G.Tont, Markov approach of adaptive task assign-ment for robotic system in non-stationary en-vironments, WSEAS Transactions on Systems,9(3), 273-282, (2010).

[29] W. Jatmiko, A. Nugraha, R. Effendi, W. Pam-buko, R. Mardian, K. Sekiyama, T. Fukuda,Localizing multiple odor sources in a dynamicenvironment based on modified niche particleswarm optimization with flow of wind, WSEASTransactions on Systems, 8(11), 1187-1196,(2009).

[30] S. Staines A., F. Neri, A Matrix TransitionOriented Net for Modeling Distributed Com-plex Computer and Communication Systems,WSEAS Transactions on Systems, 13, WSEASPress (Athens, Greece), 12-22, (2014).

[31] M. Camilleri, F. Neri, M. Papoutsidakis, AnAlgorithmic Approach to Parameter Selectionin Machine Learning using Meta-OptimizationTechniques, WSEAS Transactions on Sys-tems,13, WSEAS Press (Athens, Greece), 202-213, (2014).

[32] M. Papoutsidakis, D. Piromalis, F. Neri, M.Camilleri, Intelligent Algorithms Based onData Processing for Modular Robotic Vehi-cles Control, WSEAS Transactions on Systems,13, WSEAS Press (Athens, Greece), 242-251,(2014).


E-ISSN: 2224-2678 25 Volume 14, 2015

recent advances on locomotion mechanisms of hybrid mobile ... · recent advances on locomotion...

Documents