Allow me to introduce something I'm building in prototype and which you can see in the photo on the left there. His name is Rudebot, and he is a tricycular ROV that will, under radio control, trundle around, point at people, and insult them. Oh, he'll also have a really sardonic chuckle and he will make farty noises.


It is at this point you may ask "why?"


I reply that the world needs more gobby, sarcastic and flatulent automatons in it to make it a cheerier place. Well, it makes me a lot cheerier to see robots behaving badly.

Go Back Introduction Philosophies Driving Motors Steering a Robot


CCOG/Wheelbase rudebot

The thing about Rudebot that concerns us here is not his capacity for behaving contrary to accepted social etiquette but his systems. He's a long way from finished yet because he needs a voicebox, a boom with a pointy finger on the end of it, some bodywork and a pair of ping-pong ball eyes on wobbly stalks.

At the moment his drive systems are finished; he can go forwards and backwards at five different speeds and he can turn either way with a steering wheel that can go to five positions either side as well as being straight, those positions being shared out in a turn of 90 degrees. In other words he can either crawl or zip along, and he can turn in a tight or a wide circle. He's radio-controlled by a pair of XBees, you can see the one on Rudebot and just see the corner of the one on the transmitter which works by pushing a thumb joystick forwards and backwards for the speed and motor direction and sideways to get him to turn. I've been playing with this for a fortnight and it's great fun just as it is- I can't wait to get it insulting people. You might be able to see his brain is an Arduino Pro-Mini, and you might be able if you've got sharp eyes to pick out  what looks like three transistors on the breadboard with the Mini and the XBee. Two of them are, one isn't. What you can't see under the cage at the back is a 5V regulator. It supplies a rail at the side of the breadboard which in turn supplies the big transistor's (which is a Darlington) collector and something you can't see that's slung under the chassis which is a relay for the H-bridge, or at least the coil supply. It feeds the steering servo too that turns the small wheel at the back.  At the other side of the breadboard is a rail that supplies 9V  directly from those 6 AA batteries which feeds the Arduino, and the 3.3V regulator that supplies the Xbee which is one of those things that looks like a small transistor in a TO92 casing. The other one of those things that looks like a small transistor in a TO92 casing is in fact a small transistor in a TO92 casing and it doesn't take power from anywhere; it switches the relay causing a change in direction. The motor controller feeds a motor (well it would do, now I come to think of it) and the motor turns a multio-ratio gearbox which turns the big wheels at the front and Bob's your uncle.

Servo. Motor. XBee. Pro-Mini. Regulators. Transistors. Are you beginning to see something here? If you think Rudebot's going to need a bit of power you are absolutely right. I haven't done an endurance test yet but at the current rate I think I could run Rudebot on those six Maplin batteries for a bit short of 90 minutes before I'll need to change and that's a hopeful estimate. I can't run him on a 9V battery; I'd drain it  flat in no time. I cant use mains as he's supposed to have a range of at least 100 feet from the operator and that's a very long cable run. And I can't use C or D size batteries because of the size. I can't use four batteries to get 6V because the Arduino and the 5V regulator won't like it and so I'm stuck with 6 AA batteries for 9V. When he's finished I'm going to try rechargables at 7.2V and see what happens. I think I can get away with it but  there's always Lang's three laws of robotics:



1.A robot may exasperate a human being when, through inaction when it's supposed to be doing something, or action when it's not supposed to be doing anything, allowing a human being to come to a state of utter despair and bewilderment.


2.A robot won't obey the orders given to it by human beings if it can help it, because such orders would conflict with the First Law.


3.A robot must protect its own resistance (to doing what its maker wants it to) because  such protection does not conflict with the First or Second Law.


And of course the general law of electronics:


Any flaw in the design will not become apparent in the most stringent of testing conditions but will manifest itself immediately when the design is being presented to everybody else as soon as the most embarassing moment for the flaw to be seen occurs.


That was quite a lot of rambling even for me and so let's get to the topic in hand which is how you can power your bots. The simple answer is electricity. As the unfortunate engineers at Honda whose job it is to make an electric car that goes further than the end of the road before having to lie down and have a rest will tell you, providing juice for electric motors in great quantities from an on-board power supply is a hard task. So how can we get our bots moving on electric power?


Believe it or not there is a robot that eats. It's called ECObot III (there were two more before it and now they're developing a fourth) and it was developed by Bristol Robotics Laboratory (University of West of England and University of Bristol). Here's a picture of it: (click on the picture for BRL's website)




Well, yes, it's pretty big, very ugly and it doesn't do much except eat and move slowly about in its own little world. But don't knock it because a. it can catch its own food (really!) and b. it is the only robot in the world to have a full digestion system that creates energy from the food it digests and c. I know at least three people who are exactly the same. Mind you even they wouldn't eat what ECObot III does because what it eats is flies. It has a fly-trap at the top, and it emits artificial pheromones. Along come the flies looking for a party, but instead they fall into a pond containing a microbial inoculum and cease to be examples of Musca Domestica and instead

are digested into a nutrient rich solution of

yummy robot nosh which is pumped into forty-eight microbial fuel cells (MFCs) which are the orange bits that you can see in two rings around ECObot in the picture. Within each cell, there are microbes breaking down the food (the former flies) and ions are produced. Ions have an electrical charge. If you open it to the air you get an absorption of oxygen and electrons get exchanged across a membrane between the anode and cathode. Voila! Electricity. But not much, I'm afraid. Nevertheless it  is a remarkable achievement. The robot catches its own food, digests it, derives energy from it and then literally excretes it. Yep. Robot manure. I'm glad I don't have to clean it out. This has set scientific tails wagging the world over (although the flies are not too keen on it )  but it doesn't have any practical application in the technical field yet.. If the process can be made far more efficient and far less disgusting though, it could be that in 100 years or so we will have a robot that can feed itself and do useful things too, or a dead-fly powered battery. As a by-product the researchers have advanced the technique of 3d printing (that's what a lot of it was made with) and found  techniques in very low power circuitry which will impact the technical field more quickly.


Alright, that's an interesting diversion but we are not going to be powering our bots by liquidising any Musca Domesticae although it would be nice if we had a wasp-munching bot. I'd buy that, I hate those stripy yellow pests especially when they decide to dive-bomb my coffee and don't pull out in time. Particularly when my back is turned and, like last week, it's only in the nick of time that I notice as I'm about take a big gulp that some horrid little vespula vulgaris has decided to go for a swim in it. But I digress. In technical I am afraid we are currently at the stage where less biological means of providing electricity are prevalent. So let's look at solar power first.

On the left is a map produced by ESTELA (European Solar Thermal Electricity Association) showing the rate of insolation around the globe. Insolation is how much power in Watts the Sun delivers to a specified area of ground (square meter or kilometre) and it's good news if you live in  Western Australia or in fact anywhere in the rest of it, China, India, Pakistan, Bangladesh, various bits of the Indian Ocean, the middle of Africa, Brazil and in the United States Florida and California (perhaps that's why they wear "Daisy Dukes, bikinis on top"......I'm watching you Katie-bloody-Perry and I don't like it but at least you enunciate more precisely than Cheryl-bloody-Cole).

In Spain and Portugal it's going to be a piece of cake to get your bots working off solar panels and in the UK it's........oh. I'm afraid that if you have a coastline on the North Sea or the Baltic or your head of state is Vladimir Putin, or you live somewhere where making a cup of tea takes about three hours (yes Japan, you!) then it's not going to be a feasible thing to do. You Canadians can forget it too I'm sorry to say. Actually I'm exaggerating a bit there as we in northern latitudes do get some sunshine occasionally. About a week in August. And with luck three days in September. If it's not raining.

It is possible to get very small robots moving with small solar panels in the UK and similarly-climated countries. Here's a video of something I made using a kit from Maplin which lets you make six different things using a very small solar panel.

That kit is actually very good and at the time of writing Maplin were selling it for £10; you get lots of bits and kids love it. As an educational tool it's great because they learn construction, electrical and physics stuff without actually realising they're doing it. But you can see what the problem is. That panel is driving a very small motor, there's plenty enough torque there to drag a small and light vehicle across the floor but nothing that can do anything that needs a bit of grunt behind it. Similarly if the vehicle turns out of the direct line of the Sun then the fuel supply gets cut off and the vehicle stops. You could of course combat this by putting arrays on the front, top, back and sides of a vehicle but this will prove expensive and leave the vehicle vulnerable to all sorts of damage and it won't solve the problem if the vehicle goes where the Sun can't get, such as under a car, behind a bush or down a large pipe.  And if you are going to test vehicles going down pipes, make sure it's a pipe you can comfortably squeeze down before you do if you need to retrieve it, or you'll find the subsequent scenes somewhat embarassing as you try to hook it back out with a wide variety of unlikely objects. Same goes for bushes. And yes I am speaking from experience here.

As with digesting systems, clearly solar power is not a terribly viable option unless we can somehow store a charge. This means either a large capacitor or a battery, in which case we might as well start with a battery in the first place.


Let's talk about wind power. It's ridiculous. Enough said, so let's go on to a more realistic concept which is a dynamo. A dynamo needs a prime mover and you could of course use a small nitro engine designed for a model aircraft to do it. The most common ones are two-stroke and because of the preponderance of the American manufacturers in this market the size is expressed as the size of the bore in cubic inches and the size of 0.4 cubic inch is given as 40 and 0.61 cubic inch as 61. You won't find this used commonly as there are serious drawbacks.  Apart from the gearing, very high voltages and currents can be produced from the dynamo if driven by an internal combustion engine. The problem of vibration is also one that larger devices can handle but can be fatal to smaller ones. The fuel is highly combustible and needs safe handling and the mechanical maintenance needed to keep the engine running is highly inconvenient. Bots using an internal combustion engine can't be used indoors because of the exhaust fumes. So how about a dynamo powered by clockwork? A Bayliss radio works by slowly unwinding a spring, which turns a shaft which turns a dynamo which provides the power which feeds the detecting, discriminating and amplifying circuits to provide a sound output. Radios do not require that much in the way of power, whereas motors do.  Whilst it's possible to scale up the clockwork and the dynamo you'd have a large and weighty object to accommodate and you'd have to wind it up every few minutes. Wind up recharging is even worse. You'd be constantly spinning the charging handle on your device.


It's clear that the most suitable universal choice is the battery. The question now is: which sort? Over the page we look at a multitude of different sorts both in chemistry and performance.


Ian Lang, August 2013

Battery Concerns Scrapbots Brains Practical 1 ROV Controlling