Daily News·
Timeline·
R & D History·
 
  Research & Development History
 
 
Phase One prototypes are intended to prove a solid concept to base a design off of.

Phase Two implementation plans are the result of research based on the proposed structural design.

Phase Three assembly plans detail what kind of materials will make up what parts, and where they will all go.

Phase Four final adjustments apply kinematics results with expected test data for weight and sizes.

Phase Five metal to metal work finds every margin of error and resolves it part by part.

Phase Six testing is a procedural list of abuse that parts will undergo.



Phase One; Structural Prototypes(back to top)

Overview: Phase one prototypes are intended to provide a valid concept for a functional walking mechanism that allows a 'stand-to-walk' mobility while conserving as much energy as possible. Simple phase one prototypes were constructed using LEGO bricks, a children's toy. Thanks to the wide variety of parts available from LEGO, generating physically-viable concepts was quick, and designs were easy to modify. Plastic bricks were a logical choice to minimize the number of fully modeled concepts that would need to see development.

Experimentation eventually boiled ideas down into one of three main walking-mechanism prototypes:

Code Name: Prototype0
Common Name: Walker Mech
Description: This is a standard knee-swing, forward articulation leg system. A human being is a close comparison for bone structure. This design works off of a knee arrangement, with off-set 'muscle' girders mechanically connected to and powered by a spindle (right leg is at 0 deg, left is at 180 deg.) Due to its design, this leg system does not require a fragile ball-and-socket and thus avoids the problems associated with using that style of joint. Prototype0 uses purely circular motion from an engine source, which allows it to function directly off of a normal combustion engine.
Flaws: This mech walks well, but because of the 180 deg. offset, is always in mid-stride. This design as-is cannot stand up perfectly straight.
Analysis: Does not meet stand-to-walk criteria as-is. A secondary mechanism to adjust the mech's posture when sessile is required.

Code Name: Prototype1
Common Name: Chicken Walker
Description: This prototype uses a reverse-articulated joint similar to the chicken's. A popular example of this type of leg structure is the robot in the RoboCop movie that could not walk up stairs. Because legs in this type of arrangement can be front-mounted and significantly longer than the knee-swing arrangement of Prototye0, a mech built in this style could travel at greater speeds and work purely from hydraulics. Tip-forward and center-of-mass balance issues are also reduced with this design.
Flaws: Although the stride length of this model would be fantastic compared to other designs, too much stress is placed on the feet. Balance is also a large issue with this prototype (like it was in the RoboCop movie), the tradeoff for its tremendous stride length. Lastly, for all its speed this design is fairly inefficient for actual locomotion.
Analysis: This design as it stands is almost good enough to implement, but will need significant later-phase development before it can be trusted.

Code Name: Prototype2
Common Name: Falling Mech
Description: This mech operates in a constant state of falling, working off a principle common to orbiting. To walk, one leg is both pneumatically shortened and raised up (stepping forward). A large mass then slides to the now-falling side, speeding it up. The leg contacts and is then re-heightened as the opposite leg begins its cycle and the weight slides in the other direction.
Flaws: This design requires a limb with internal pneumatics, placing excessive stress on the internal structure and generating a weak point. The limbs will no longer be solid, changing the appreciable physics a bit.
Analysis: Efficiency and structural problems require later development. An energy-return system in particular needs to be implemented.

Final Results:
None of the immediate results from phase one produced a walking mechanism that was ready to go. While there were mixed results from each concept, the merits of a complex control system, the mechanical simplicity, and the potential of Prototype0 made it the leading contender. Development into this design included exploring the possibility of either a secondary rotation mechanism, or an extendable/contractible leg segment to alter the leg alignment when the mech was not mobile.

Phase Two; Implementation Plans(back to top)

Overview: We have since completed research and development in several areas that presented themselves as issues. Each is briefly outlined here, with hyperlinks to the full brief (if available).

Structural Components
Initially we had opted for a very bottom heavy design using 1 inch thick and 12 inch wide steel I-beams. It seemed cost effective, and with the lower center of gravity, easier to turn and stabilize. The legs would also be the part receiving the most punishment in all aspects of movement. This over-preparation served us well in the long run. We were able to modify the design because of its simplicity, and had plenty of weight that could be removed by switching to lighter weight materials and using I-beams modified to take more stress while undergoing a minimal amount of plastic deformation.

Power Source
Our initial power source was to be a 396 cubic inch ford engine with a race cam. Due to the superior engineering of the General Motor's small block, and for weight vs. performance issues, we opted for a 350, which could be stroked out to a 383 for an increase in torque. Additional horsepower may be attained via a supercharging unit or turbochargers.
Further power to weight considerations may see a custom rotary engine in place in the long run.

Balance
Balance will be measured by the use of piezoelectric gyroscopes, assigned 3 to a unit, and will provide position information up to the millisecond to the Piloting Control Interface. Additional feedback data is returned by an RVU on each joint and plane of motion in the machine. This allows the machine to manage its balance autonomously, like a human, by using its body.

Energy Conservation
After the aerodynamic calculations and energy requirements were revisited, we were happy to find that our reductions in overall frame size, and weight have accounted to just under 50% of the original power requirements. This gives us a 22% margin over the power:weight ratio line that has previously prevented the use of large scale walking vehicles.

Phase Three; Assembly Plans(back to top)

Overview: It was expected that we would face issues with our selected components once we started assembly. Development progress in each area is outlined here.

Energy Conservation
Our own version of Juice, a physics simulation engine by MPS's Nate Waddoups, provided us with a test-bed for real world designs. This allowed us to develop an uncomplicated setup without having to assemble physical prototypes. The most important step in the development of the kinematics was ensuring that we would have an efficient and practical way to facilitate mobility of the unit. Using straightforward solid state techniques, we have engineered a design which minimizes non-harmonic movement, and focuses on mass and inertia centralization. A series of passive pneumatics and active pneumatics and hydraulics will be implemented to create a comprehensive energy conservation and return system.

Propulsion Systems
MPS's early established trend of relying on proven, existing technology led us to the Chevy family of engines and Porsche family of automotive components. By adopting automotive level components from existing manufacturers with a reputation for quality engineering and components, MPS has reduced the development time of the machine in all aspects of the mechanical assembly.

Materials and Assembly
Probably the most important aspect of the machine's structural elements is the materials used. By selecting a range of materials that were best suited to use, MPS was able to compact the materials list to a few readily accessible materials with amazing properties. The steel selected by MPS can be heat-treated to become up to 33% stronger than titanium, is relatively inexpensive compared to exotic metals, and has weld-ability characteristics on par with mild steel. Any non-facing machined parts can be manufactured from billet pieces of our choice of aluminum. Minimum deformation under machining heat, as well as as tensile strength greater than mild steel made this inexpensive and lightweight alloy the perfect alternative to a heavier steel.

Piloting Control Interface: Software
After brief proof-of-concept programming, a GPL released 3d engine was tested for use. This provided us with a significant testing ground for technology, without the labor and development time required to implement such a testbed. We have now developed a failover-capable communications infrastructure that relies on our own version of neural network enabled branch prediction and process resolution control software. Even though our software is developed in-house, MPS relies on existing standards and an open architecture. Our base operating system is a stripped down version of Linux. Constructs and visual elements use other industry standard components such as the SDL visual library and OpenGL to expedite development and increase CPU efficiency.

Piloting Control Interface: Hardware
We purchased a couple of i-Openers, which were modified to suit our needs. iOpeners are a lightweight internet appliance manufactured by NetPliance and were created from commonly available laptop computer parts. iOpeners utilizes the 'x86' or Intel/IBM Compatible architecture. The initial units were underpowered, and a few modifications were completed:
- Debian Linux was selected from the list of candidates and was installed.
- An increase in RAM was done.
- Ethernet networking capabilities were added via a tiny USB card.
- A 4.3 Gigabyte laptop harddrive was added.
- An external CPU fan and heatsink was added.
- We intend to upgrade the 180Mhz winchip processors as well.
The concept of using laptop components proved sound. Lightweight computer components further reduced the expected weight limits required for the cockpit. The introduction of Mini- and Micro-ITX formfactor computers reduced size and energy requirements. A short step of migration to solid-state storage components, and ruggedized hardware, coupled with industrial vehicle communication protocols results in our final hardware configuration.

Phase Four; Final Adjustments(back to top)


Engine Options and Specifications
Base Engine, Mods.DisplWeightTorqueMax HPOctane
454HO GM Crate Motor
Perf. Dist. "DUI" HEI distributor
Dynomax Headers #86110
825cfm Demon Carb
Edelbrock Victor Jr. Intake
Vortech YS-trim Supercharger
Crane hyd roller cam #CRN-168761
Crane hyd roller lifters #CRN-16535-16
Crane valvesprings #CRN-99896-16

454cuin. (7.4l)700lbs (320kg)706@5,5779@6,191
Chevy Vortec Smallblock
750cfm Edelbrock Performer Carb
Stock GM HEI Distributor
1" Carb spacer
1.6:1 Alum Rockers

350cuin (5.7l)551lbs (250kg)452@4,4442@5,991
Caterpillar Diesel Engine; 3406E
Turbocharger
Air-to-air aftercooler
ECM Monitoring System
Flywheel and housing

893cuin (15.8l)2 897lbs
(1 317kg)		1 850@1,2575@2,1Diesel
Caterpillar Diesel Engine; 3116
Torsional Vibration Dampner

403cuin (6.6l)1 195lbs
(543kg)		890@1,2275@2,4Diesel

More information is in this technical document: MPS_Power_Plant_Choice_Discussion_1.0_FINAL.doc

The current powerplant for the Mk I prototype is a fully rebuilt Cadillac 425 engine with moderate cam work to improve torque and high end horsepower. This may also include the use of a supercharger and/or nitrous injection, both of which are already acquired and awaiting use.

Later marks will use a twin engine alum block supercharged Chevy 454 design. Propane injection may be used increase the octane value of the fuel to prevent predetonation inside the combustion chambers under very high boost.

Kinematisch Masse
MasseschreibungMaschineTest A%teil
Beinhoehe340cm85cm400%
Pelvishoehe420cm105cm400%

StridetypLaenge fur Test A%teil zum BeimhoeheGeschaetzt fur Maschine
Weg80cm90%320cm
Laufen215cm252%860
Maximum268cm315%1075

Bein RPMDehrenderGeschaetzt Masse fur Maximum Laufen
157.5~9.7kmph (6 MPH)
3015~19.4kmph (12 MPH)
6030~38.7kmph (24 MPH)
12060~77.4kmph (48 MPH)
18090~116kmph (72 MPH)
240120~155kmph (96 MPH)

Calculation notes:
There is some additional length (180cm / 6') that can be applied in forward rotation. This additional length comes form the exaggerated pelvic sway present in the Mecha. Test Subject A represents a 30 cm (12") pelvic dispersion width, but uses only 4cm (2") in forward rotation. This represents a material gain of about 16% of the given width. The Mecha would ideally take greater advantage of said distance, optimally 100cm more beyond the 30cm expected. The aforementioned length would increase the overall stride length some 100 cm (40"), representing a possible 9% increase in nominal speed.

Final Dimensions
Total Height: 22'
    Torso 8'
    Walking platform 12'
    Head 2'
Total Width: 8' plus armature
Total Depth: 5' including entry hatch hinges and access ladder.

Phase Five; Metal to Metal(back to top)

Mech Overview
    Physical Spec - less than 3,000lbs, under 24' tall, under 10' wide, 8' wide torso, capable of pilots between 5' and 6'6" and 80-400lbs.
    Concepts - Simple, easy to assemble, acquire and maintain, industry standards on hardware, software, protocols, equipment. Existing, proven technology.
    Control - Internal pilot required, no remote operation except basic troubleshooting and loading procedures, heavy use of software neural networks, and either simple, clear-cut custom interface or existing interface devices.

Structural Components
    Frame Overview - Cromoly roll-cage arrangement with a robust turning and locomotive mechanism.
    Head - Simple box with breakaway protection, having sensor equipment.
    Torso - Roll caged superstructure with Lexan sheeting, and space for pilot and  racks of hardware.
    Legs, Upper - Robust and hefty. Capable of taking the step load with no deflection.
    Knee - Simple and robust, capable of locking and deceleration.
    Legs, Lower - Lighter and capable of cushioning a step.
    Ankles - [[Proprietary Data]]
    Feet - Cromoly triangular superstructure with aluminum pressure plates, and a passive toe.
    Arms - Purchased assembly to spec.

Main Subsystems
Propulsion and Support Systems
    Internal Combustion Engine - Cadillac 425 cubic inch 7.0 liter engine with slight modifications.
        Venting for ICE - Heat wrapped custom exhaust headers with 
        Expel Heat - Dual radiators with auto-switching twin electric fans and an oil cooler.
        Muffle Sound - sound/heat insulation with intelligent acoustic design
Positioning and Control
    Passive/Active Pneumatic Systems
    Reactive Hydraulics Systems - Standard aviation hydraulics.
Mechanical Walking Mechanism - Custom machined parts attached to standard drivetrain
    Power Transmission - Robust automotive transmissions, differential, and 4x4 geardowns.
    Clutching Mechanisms - [[Proprietary Data]]
Control Interface - [[Proprietary Data]]
    Pilot Control Interface - Proprietary coded interface with open architecture hooks.
        Display Screens - Industry standard 
        Physical Interaction Devices - Industry standard interface hardware
Systems Interface - Ruggedized, failover capable industrial communications network
    Control Mechanism Interface - Neural Network enabled autonomous machine management
    Inertial Balancing Systems - Neural network controlled synchronization matrix
        Gyroscopic Balance System - Inertia Reference System embedded feedback hardware
        Inertial Assistance Mechanism - Redundant Vector Unit embedded feedback hardware
Pilot Restraint System - Passive and active systems to improve the physical pilot experience
    Head Restraints - Protect pilot's head and neck from fatigue or injury.
    Pilot's Harness - Standard 5 point high speed sport racing harness.
    Active Impact Prediction - prevents pilot injury by prediction, compare to airbag technology.
        
Phase Six; Procedural Testing(back to top)

Computer modeled engineering data coupled with real-world stress testing ensure a delivered product that exceeds expectations. This creates the perfect marriage between intelligent design and application testing. Parts are modeled for efficiency, to be light weight, and robust. Problem areas are cited and tested. Real world assembly then reflects on the next generation of parts to expedite manufacture and installation of individual components.  

Actual test data is proprietary and is not publicly released.