ROBOT GIANT ANT FILM SPECIAL EFFECTS MODELS ANIMATRONICS MOVIE INSECTRONICS WRITTEN BY JAMESON HUNTER LIMITED ORIGINAL STORIES FROM THE SCOTTISH AUTHOR JIMMY WATSON AND THE MAGIC DINOBOT A CHRISTMAS WISH COMES TRUE

The CAD drawing above shows the head and jaws (see the full size pattern in gold below), head to main frame joint and leg joints (no legs). The motor/pump and timing shaft are fitted above the main frame shown in blue. An autonomous version of this robot would make a superb artificial soldier.

Giant insect - body casing/ artwork for the robot mechanics described below. The human skull in its jaws is full size - to give scale. The head shape is easily identified as that in the CAD drawing to the left. The large hump in the main body segment is ideal to house the hydraulic pump and powerful Lynch electric drive motor shown further down on this page.

SECTASAURUS - New for 2016, Jameson Hunter is now developing the giant (bulldog) ant based special effects animatronic for this potential cult sc-fi story. Bluebird Marine Systems have been asked to look at re-making the mechanics that went missing many years ago. The object is to produce a few publicity articles and maybe give some aspiring film producers a chance to make a pilot, possibly with the help of crowd funding.

Back in 1982, an English inventor was working on a robot inspired from the form of an ant. He'd been studying ants for some years, keeping wood ants (formica rufus) in a formicarium in his workshop. Ants are capable of lifting 200 times their body mass, due in part to their exoskeletal structure, where the hard body shell carries the structural loads, rather than an internal skeleton and soft outer body tissue. The body of an insect is made of a natural form of plastic, a derivative of keratin. That also means that as an engineering form, the ant design is pure genius molded from millions of years of evolution.

Insects and Arachnids are armed some incredible weapons, that make mammals look positively backward. Even reptiles have piercing fangs and a poison bite. We only have deadly wits. Okay, we don't need to stun our prey so our children have a ready food supply. But the taxman can sting just as hard with a piece of paper and suck the life out of us. Insects have adapted to just about every niche in their small world, with superbly inventive remedies.

Eventually, a film company took on the inventor's idea, and these are some of the original hand drawings produced in connection with that project, salvaged from the archives. Remember that in those days computers and CAD design was a pipedream unless you were a billionaire:

ELECTRICAL CIRCUIT - There are 18 valves for the legs of this robot, each with three positions: Forward, Lock, Back - so, 54 states for each valve to be in. No feedback loops are included in this basic configuration, which should be seen side by side with the hydraulic circuit below.

The circuit diagram above is the electrical switching sequence that energizes 18 solenoid operated three-way hydraulic control valves for a possible 54 valve positions, including the locked positions. The firing order was worked out by studying real ants, and watching their walking movement. Three banks of brushes fed from rotary contacts on a hexagonal shaft. Each bank had a different timing angle that gave 1. Forward and backwards, 2. Left forward and back and 3. Right forward and back. Switching electrically between banks provided directional steering control.

Today, the firing sequence would be digitally programmed and operated by an onboard computer, triggering transistor amplified gate switches. A very realistic walking gait could be achieved as exemplified by the Youtube videos of robots using electric-motor servos at the foot of this page. An interesting mechanical chain/crank drive version from Spain is also shown, which although extremely simple in design, appears to walk reasonably well.

The relative simplicity of this design is a key factor. Changing from a hexagonal shaft to a decagonal shaft would increase variations. But, changing to software triggered valves with feedback from the limb concerned, to include ground contact, temperature, humidity, etc, could mimic a life form and safeguard the robot from getting a leg trapped in soft mud, etc.

The illustration above shows the brush holders, rotary contacts and hexagonal shaft drive of a pre-digital-age mechanical system, where the hexagonal shaft is live and the rotary contacts are slotted onto the shaft into the desired position according to the sequence you want. When you have an idea, draw it. Get it down on paper, or you will forget it. The timing is not unlike that in older cars for the distributor of the ignition system. Early washing machines also had rotary timing switches in stacks for the wash and rinse cycles - now digital programs. The left and right legs can be treated as left and right tracks as if it were a tracked vehicle (tank), for simplification.

WALKING CYCLE - VALVE POSITIONS FOR FORWARD, LEFT & RIGHT .
.	.	.	.
STATION	FORWARD	LEFT	RIGHT
.	.	.	.
A	01-48-54	01-47-54	02-48-54
B	31-34-24	31-34-24	31-34-24
C	37-12-16	38-12-16	37-11-16
D	04-07-51	04-07-51	04-07-51
E	19-30-36	19-29-36	20-30-36
F	15-40-43	15-40-43	15-40-43
G	46-03-09	47-03-09	46-02-09
H	33-22-25	33-22-25	33-22-25
I	10-39-45	10-38-45	11-39-45
J	06-49-52	06-49-52	06-49-52
K	28-21-27	29-21-27	28-20-27
L	42-13-18	42-13-18	42-13-18

HYDRAULIC CIRCUIT - There are 18 valves shown in the center of this diagram and 36 pistons to operate the legs of this robot. The head and tail pistons are also shown with 5 more 0.25" spool valves. The above diagram should be seen side by side with the electrical circuit above.

The hydraulic circuit diagram above shows the head, leg and tail rams. The solenoid operated valves for the legs are arranged on the body main frame in two banks, with an oil reservoir, pump and pressure vessel. The walking cycle leg activation order is: 1-5-3-6-2-4. This sequence equals one x walk cycle.

Taking the forward movement of the robot as 36" (3.0 feet) per walk cycle, we can say that one revolution per second will be 2.0 miles per hour, where 1 mph conveniently = 1.466 ft/sec (17.59"). We won't worry about absolute accuracy here. We just need to have a ball park overview to get a handle on what is possible when designing conceptually.

The hydraulic fluid flow per walk cycle works out to 452 centiliters per second, which equates to 1.357 liters per second @ 6 miles per hour, 4.071 l/s @ 18mph and so on. The leg pistons as shown in the diagrams on this page have a bore and stroke of 20 millimeters = 12.568cc. There are 36 pistons. So one walk cycle = 12.568cc x 36 = .452 liters. With this timing system, the walking speed of the robot is governed by varying the motor revolutions. If using a computer to time the valves, more versatility/agility is built in.

Plan and side views of a 10kW (13.4hp) 60v dc Lynch motor, mated to a hydraulic high pressure vane pump (from auto power steering). This setup is more than capable of powering our robot soldier ant to very high speeds, without straining the motor. Faster in fact than a human can run flat out. You'd have to go some to match a Cheetah = 16 liters a second.

BODYWORK - Click on the picture above to see how to make giant insect body parts quickly and cheaply. This is an educational tutorial for anyone interested in making animatronic displays.

To make this robot effective when engaging humans, a 21 mph running speed would require a fluid flow rate of 4.746 liters per second. Look out Usain Bolt. This figure is attainable, which in turn would need a powerful electric motor coupled to a power steering vane type pump, or a similar high pressure vane pump to those used on fishing vessels for hauling in the nets.

The jaw closing pressure was calculated as 0.78 tons and .42 seconds from full open to full closed. This of course can be changed with the area of the pistons and available volume from the pressure accumulator chamber, which could see 2 or more tons pressure in less time, if that was a priority. It wasn't. More likely, the military would prefer guns, the police, stun weapons and civilian operators (if any), surveying equipment.

Above, conceptualizing a design by drawing the workings. Two proposals for a ball joint with 4 hydraulic rams placed vertically, operating via cables, and ball joint with swash plates. In the end the version below seemed more practical - a natural development - but you have to start somewhere. The inventive step here was the gimbal to replace the vagaries of a ball joint, which is also free to rotate. Sometimes of course, rotation is a desirable feature - as in our hips and neck joints.

Blueprints for the (final) compact gimbal based "unit" joint, suitable for the legs and head of this particular robot and many other applications when used in conjunction with the tri-function spool valves. In any production version, positional feedback should be included.

The main leg joint (shown above) is a clever hydraulic piston cluster above and below a gimbal type central swivel joint. The single acting pistons work against each other, pushing flat machined swash plates that are oil lubricated. Single acting pistons are more reliable that dual acting rams, because there is no shaft seal, only a piston seal. This arrangement relies on precision control via a two-way hydraulic valve, which must be balanced and capable of dealing with the high flow rates that running will require.

In order to work correctly, the inventor considered the merits of several types of hydraulic control valves - as seen below. It was imperative that the valve could be set in a position to lock a piston, so freezing the movement of the robot, rather than see a gradual collapse. For this reason, three solenoid windings are needed. Once again, in the digital age, programming would take care of this, reducing production costs considerably. The point here though is that this system was designed between 1980 and 1984 to give a robot capable of lifting humans, or giving them rides, etc. The real thing. One can imagine the uses for such a robot in combat. Armour it and give it a weapon with laser ranging, and you'd have a fearless, disposable infantryman. Food for thought and surprising really, that we don't have mechanical soldiers like this today for hostile regions like Afghanistan.

The rotary valve design and rotary solenoid, a bit like a stepper motor, with exploded view and assembled cutaway view. The rotary magnetic operation may provide more torque than the spool version below. - but the spool version (below) is shown on the CAD robot drawing above. Note the design to minimise fluid drag, so heat build up.

Two linear valve designs, spool and flat bar gate (above and below). Interesting how seals could be eliminated. The 1/4" spool version on this drawing from 1983 is seen in the CAD drawings above, nine units per side of the main frame. Three solenoid windings provide pressure/lock/release flow positions - essential for complex movements and economical operation.

We are unable to publish any of the original working drawings because the paper they are drawn on full size (called lofting) is something like 10 x 10 feet. But we now have CAD and have input some information to share with you, in the drawings published, so that students of robot design can appreciate the chassis configuration and design practicalities which is very much the same for a robot ant as for a robot human, save the balance issue, which is easily solved with an appropriate sensing arrangement and balance control feedback loops.

A light projector was used to obtain superbly accurate dimensions from a specimen Bulldog Ant - see below. This system is suitable for application to any insect.

It is a pity that the design is not exoskeletal (monocoque). We imagine that this would be extremely complex design wise, but much simpler for the construction.

With a design like this, mated to computer controls, any autonomous variant will be a truly awesome machine with limitless possibilities.

This was the specimen that changed the look and size of the robot ant development. The bulldog ant from Australia is a particularly interesting animal. It can be a solitary hunter. It can paralyze a human and it has stereoscopic vision. Click on the picture to see more about the making of the protective showcase.

NOVEMBER 2015 - The above artwork is with Bluebird Marine Systems, who are now looking to make the missing parts and re-assemble the giant animatronic ant for Jameson Hunter Ltd. Does that mean we might see a script of film anytime soon?.

Should film or special effects companies, natural history museums or other events organisations be interested in such a service, to include production of the artwork to a scale of your choosing, please contact us and we will put you in touch with an engineering and molding team who can provide a quote for you.

NOTE: These copyright designs are now the intellectual property of Blueplanet Universal Holdings Ltd, to be used for publicity (marketing) and special effects for the British science fiction film called Sectasaur.

A-Pod is an ant inspired hexapod robot with a 2 DOF abdomen (tail), a 3 DOF head with large mandibles. 6 legs with 3 DOF each. Total 25 servos. This video demonstrates body movement and mandible control. The designer still has to make some mechanical improvements to the legs (explaining the small amount of walking). The robot is remotely controlled with a custom 2,4 GHz RC transmitter. The Basic Atom Pro 28 is/are used as the main micro-controller(s).

Visit www.basicmicro.com for more info about the BAP28.

FireAnt is based on Zenta's A-POD and made from all aluminum anodized brackets. The kit includes the new digital HV220 robotic servos with position and force feedback. This gives FireAnt life like movement and the ability to sense its environment. It can sense the force the pincher's are applying with no additional sensors. It can also feel when a leg is touching something in both down and forward directions for autonomous terrain adaptation.

Kondo Robot announced a new, low cost, hexapod robot design with surprising flexibility, performance, and customizability. The robots unique leg design incorporates springs and linkages only requiring two servos per leg while improving the hexapod's ability to deal with uneven terrain and obstacles. The robots base frames can be easily fabricated or modified by builders to realize custom configurations. For more information visit Robots Dreams at:

A small autonomous agile robot with an on-board neurologically-based control system. Developed by engineering professor Roger Quinn and his team at Case Western Reserve University. Video presented at IEEE IROS 2009. Read more robot news at:

There are an estimated 20,000 species of ants in the world. Ant colonies have structured social system, with different castes - worker, soldier, queen and drone - all of which carry out specific tasks for the colony. Ant colonies are sometimes referred to as "superorganisms" because they appear to operate as a single entity.

Scientists in the US have built and tested robotic ants that they say behave just like a real ant colony. The robots do not resemble their insect counterparts; they are tiny cubes equipped with two watch motors to power the wheels that enable them to move. But their collective behaviour is remarkably ant-like.

By being programmed simply to move forward toward a target and avoid obstacles, the robot colony finds the fastest way through a network or maze. The secret, the researchers report in the open access journal Plos Computational Biology, is in their ability to take cues from one another - just like an insect swarm.

"Each individual robot is pretty dumb," said Simon Garnier from the New Jersey Institute of Technology, lead researcher on the study. "They have very limited memory and limited processing power."

"By themselves, each robot would just move around randomly and get lost... but [they] are able to work together and communicate."

This is because, like ants, the robots leave a trail that the others follow; while ants leave a trail of chemicals - or pheromones - that their nest mates are able to sniff out, the robots leave a trail of light. To achieve this, the researchers set up a camera to track the path of each robot. A projector connected to the camera then produced a spot of light at regular intervals along their route, leaving a "breadcrumb trail" of light that got brighter every time another robot tracked over the same path.

Dr Paul Graham - University of Sussex: “ You don't need something as complex as choice to get some of the behaviour you see in ants”

Dr Garnier explained: "[The robots each] have two antennae on top, which are light sensors. If more light falls on their left sensor they turn left, and if more light falls on the right sensor, they turn right." "It's exactly the same mechanism as ants."

The researcher explained how both the robots and ants worked together, describing their navigation skills as a "positive feedback loop". "If there are two possible paths from A to B and one is twice as long as [the other], at the beginning, the ants [or] robots start using each path equally.

"Because ants taking the shorter path travel faster, the amount of pheromone (or light) deposited on that path grows faster, so more ants use that path."

There are many other research and engineering projects that take inspiration from nature to solve problems or design robots, as Dr Paul Graham, a biologist from the University of Sussex, explained. "The classic example," he said, "is the way in which we design information networks to move packets of data around. "Ants don't have someone in charge telling them where to go, so you can [mimic this].

For instance - in a complex network, there may be a junction with different possible routes that packets [of data] could take. Packets would leave messages for each other at the junction to give information about which routes were quick." This, he explained, is the basis of an algorithm called ant colony optimisation which has already been used in telecoms networks.

And although Dr Graham doesn't see an immediate practical use for these particular robotic insects he says the study demonstrates an important and interesting piece of biology. "Lots of animal behaviour gets described using words like 'choice'. "This shows that you don't need something as complex as choice to get some of the behaviour you see in ants. "And these things look pretty cool, too."

Dr Graham Biography: During a Psychology degree I became interested in Artificial Intelligence, specifically the prospect of capturing aspects of biological intelligence by mimicking insects. This led to a PhD at Sussex with Prof Tom Collett, where we have studied spatial cognition in ants. Following my DPhil I completed BBSRC, Leverhulme and RCUK postdocs before becoming a Reader in 2012. My major research involves studies of how insects navigate around their environment. Insects are capable of amazing feats of navigation, even though they have limited neural resources. Understanding how these small creatures navigate will help us better understand how the behaviour and cognition of an animal is constrained by the environment within which it operates. I am also interested in the application of insect research for the development of autonomous robots.

Role: Reader, Evolution Behaviour and Environment, School of Life Sciences