Can you build an auto turret with Victorian-era technology?

Question

Auto turrets:

When you think of auto-turrets you probably imagine something along the lines of: high tech or state-of-the-art weapon systems. You'd be right as such technology is heavily computerized and relies on computer chips processing visual information to determine wether or not to pull a trigger. In contrast: the motors and hydraulics that move the gun are fairly simple.

The goal of auto-turrets is almost always defensive. Guard an area and shoot anything that moves. They are advantageous because they don't doubt, miss or tire. Instead having soldiers patrol an area in shifts day in and day out, you can just have a turret stand there instead. There's obviously still the risk of an ally getting shot but if you place a turret somewhere you probably don't want anyone approaching it regardless. Advanced programming solves that issue though.

I'm looking for a more primitive model of auto turret.

Why Victorian specifically? Well it's because the device relies on sight, something impossible to mechanize unless you have access to electronics. You could of course make a rope activated ballistic turret, but that would be more of a boobytrap. The main purpose of an automatic turret is to take aim without human aid. Placing tiles on the ground that allow a crossbow to aim is not only excessive but impractical. A camera equipped turret you place in a hallway would be much more effective. So next case in point: photoelectric devices.

Does the Victorian-era offer the necessary materials for photoelectric sensors?

This is in essence my question right there. I want to make a primitive gun on a tripod that moves when the light source is perturbed. It might misfire if too sensitive so the mechanism will have to be tuned accordingly. It also still requires a steady supply of power.

The intentional weaknesses of the device are:

• It doesn't fire at stationary objects. The mechanisms only trigger when a light source if perturbed. Not if there is contrast. So if you stand perfectly still it won't shoot you anymore.
• If power runs out it doesn't do anything. Obviously, but you can't tell at first glance. You'll have to throw a decently large object to trigger it. You can waste its ammo this way but the gunshots will alert everyone to your presence (now all of China knows you are here).
• Standing too close to it blinds it. If you fill the frame there's no perturbation, unless you wear patterned clothing. You can also cover up the sensor or place something in front of it.
• You're safe if you go behind, under or above it. If you're guarding a hallway then maybe getting behind the turret isn't such a big deal. However, it can rotate 360 degrees then the owner might want to stay underneath the thing to not accidentally get shot.

Answer 1

Does the Victorian-era offer the necessary materials for photoelectric sensors?

It looks like the photolectric effect was discovered in the Victorian era

Johann Elster (1854–1920) and Hans Geitel (1855–1923), students in Heidelberg, investigated the effects produced by light on electrified bodies and developed the first practical photoelectric cells that could be used to measure the intensity of light. They arranged metals with respect to their power of discharging negative electricity: rubidium, potassium, alloy of potassium and sodium, sodium, lithium, magnesium, thallium and zinc; for copper, platinum, lead, iron, cadmium, carbon, and mercury the effects with ordinary light were too small to be measurable. The order of the metals for this effect was the same as in Volta's series for contact-electricity, the most electropositive metals giving the largest photo-electric effect.

Material wise, it's therefore a positive answer: they had the materials needed for it.

Answer 2

Use biological eyeballs.

Specifically, train animals. During WWII there was an effort to make pigeons the guidance system for missiles.

It wouldn't have to be pigeons of course. Any animal with decent eyesight (or even superhuman eyesight... this thing could work in the dark!) should be sufficient. Most are trainable. Of course, the animals would have to work in shifts, so there's some maintenance where the pair or triplet are switched out after 8-12 hours and replaced with a fresh "crew". There would be details too, like how to best keep them from snoozing without making them so uncomfortable that their performance is degraded. But you get so much more for it... they can likely be taught to distinguish between uniforms, just as a for instance. They can be taught to target only humans, and not other animals. To not shoot at leaves wafting through the air on a stiff breeze.

And other than the animal training (which to one degree or another already existed in the Victorian era), everything is mechanical and well within their capability to produce to sufficient tolerances.

Answer 3

Use contact plates - switches activated by stepping on them and/or capacitance detection for people touching fence wires.

Primitive switching can detect which detector tripped and aim the gun at it.

Answer 4

A simple visible photosensor isn't enough for aiming, and even a photodiode array (which would be hard or even impossible to make ) would be of limited use. Infrared would offer more options but anything that would help is too recent.

On the other hand acoustic location was used in WWI which is only a little late. As I commented under another answer, sensing is only part of the story. To drive aiming motors with a small signal you need an amplifier (as well as a way of comparing inputs if you're using the difference between two signals as may systems, optical or acoustic, would). I'd set a date of 1906, based on developments in valves (vacuum tubes). You'd also need a form of fire-control computer, which could be mechanical.

All in all, I reckon early 20th century is your limit for an approach of sense-aim-shoot, unless you bring forward a few real-world inventions by a few decades.

Answer 5

You could do something like this mechanical by placing tiles in a grid on the ground in the range of the tower. Each of these tiles could be connected hydraulically or mechanically to a mechanical computer that calculates aim and triggers it.

It could be made that each tile only triggers one or few shots, thus once the attacker remains stationary the turret will not continue to engage with them.

You can find a good introduction to mechanical computers that were used on Navy ships before the advent of electronics to drive turrets here: https://www.youtube.com/watch?v=gwf5mAlI7Ug

Answer 6

I know you specifically were going for purely optical (human visible light range), but - the first thermographic sensors were invented before the Victorian Era: https://en.wikipedia.org/wiki/Thermographic_camera#Discovery_and_research_of_infrared_radiation (this is the relevant citation: https://www.sciencedirect.com/science/article/abs/pii/0020089162900234 ) claims that in 1833 a sensor existed that could detect a human 10 meters away, and "By 1901, it [a slightly different device, the bolometer] could detect radiation from a cow from 400 metres away and was sensitive to differences in temperature of one hundred thousandths (0.00001 C) of a degree Celsius".

So with some tuning, one could definitely use one of those devices as your sensor - tune it to human body temperature, tune the range of sensors etc. to match your criteria (only target on large changes where, for example, 50% of the sensor view is at human body temperature).

Answer 7

It's going to be expensive but it could be done at any point after the development of the photoresistor. It won't be a perfect match for what you're after, though.

The lack of amplifiers is going to be a problem, severely limiting your options. You have some optics that focus the image onto a sensor. Each pixels of the sensor is a photoresistor. This is one half of a voltage divider. The junction point is connected to a small, sensitive bidirectional relay to large capacitor. A change in the resistance of the photoresistor changes the voltage at the divider, the capacitor will adjust causing a current flow only when there is a change in light levels. A very slow change (the normal day/night cycle) will not cause enough flow to trip it.

Both sides of this relay are wired together, you only care about current flow, not direction. This current energizes a larger relay which latches (when it closes it feeds voltage back to keep it closed) and a separate circuit in the relay connects a point on a potentiometer. The actual gun is also connected to an identical potentiometer. These two points are connected to another bidirectional relay, one side causes the turret to move one way, the other causes it to move the other. When the voltages match the gun is pointed where the sensor said there was something and the relay opens again. Meanwhile a capacitor has been charging, when it's charged above a reference voltage (another voltage divider) current flows through a relay and pulls the trigger. When the gun fires it also closes a circuit which temporarily cuts off the power to the firing part of the circuit, thus releasing the latched pixel(s). This circuit is also current-limited, if too many pixels trip the draw is too heavy, the voltage goes too low and the system resets without firing. This is a necessary safety circuit because otherwise it's going to think every cloud going across the sun is a bad guy. (But it also means the bad guy can slip something across the lens.)

This is long before electrolytic capacitors and these capacitors need to be big--which means they're bulky and expensive. I don't know how sensitive the relays can be so I can't come up with any numbers.