Der Verlorne Haufen: Chapter 3

Submersibles Deliver

River mouths and channels, where muddy brackish river water flows into the rolling sea, often experience turbulence at times in the tidal cycle. Standing waves are normally unremarkable. Yet on this island, there is a strange hydraulic phenomenon: regular waves moving upstream at walking pace. Keen eyes peering into the darkness might see something looking like the swollen fin of a dolphin breaking the surface of the static wave. Sensitive ears might hear an underwater hum intensify, changing pitch as electric motors increase thrust to counter the river’s outflow. 

In the distant Darwin home control room, the tension amongst the clustered teams of remote pilots and drivers is palpable. They all look up onto the big screen to where the map display outlines the sweep of the river, the mangroves, myriad muddy islands and above all, the curved underwater deeper gullies that are formed by tidal inflows and outflows. A concealed approach. The depth readings being transmitted from the cargo-carrying submersibles match the virtual charts, yet there could still be undetected obstacles in the channel. The LIDAR subsurface drone scans are impressive, but they do not detect submerged wires and cables. 

On the main screen a blue flashing dot representing a submersible head upriver moving slowly within the lines indicating the deeper channels. As minutes pass other, more subdued dots appear around it and become a cluster. A cascade of data spills down the side of the screen, indicating that the sensors on a watching drone somewhere high overhead and far offshore are now detecting the submersibles as they start to break the water’s surface. The operators begin to return their attention to their own screens, confirming the signal strength of their individual satellite links. They wait for their local screens to show images from the cameras mounted on the submersibles’ retractable masts. The pilot operators are poised to take control from the platform AI. Initially they can see very little. Their visual-range light intensified pictures reveal only the outline of the mangrove foliage ahead. After a minute or two the lenses of the thermal cameras on the submersibles have dried sufficiently to add heat images to the composite. As the images appearing on the pilots’ screens become clearer, there are no visible threats, neither person nor machine. 10 

Our imaginary observer with eyesight able to pierce the darkness might now see more fins breaking the surface, and for a brief moment think they were witnessing the bizarre spectacle of pilot whales on course to beach themselves. Any illusion of a natural phenomenon is broken by the surging growl of white noise as the first wave of the 15 m long, 40 ton Sublanders

open their track covers and the scooped rubber tracks begin to move. A few hundred metres from the beach the amphibious platforms slow again. They adopt a loose extended line formation and as their ballast pumps cut in, their decks begin to break the surface. From each, a glycol smoke plume rises and billows across the water towards the landing beach. The upriver approach not only provides an unexpected approach from the inland side of the town and airbase it also means that the offshore seabreeze works in their favour. Tracks now move faster and thrash the water. Hidden from human viewers by both the smoke and darkness, one after the other, the whale-like metallic bodies start to emerge from the water. They wallow for a bit as their tracks scrape at the riverbed. As they gain purchase on the sand they steadily rise up the beach. 

In Darwin, one of the pilots sees the images from her submersible swaying wildly. The AI is evidently not recognising a sunken obstruction, so she switches off the thrusters and reverses the track direction. Once the submersible platform has backed clear of the old concrete water pipe it had been straddling, the pilot moves it a few metres to the side and then forwards again onto the beach. If manual control is required to cross or avoid an obstacle that the vehicles algorithms can’t cope with, the standard procedure is to maintain the human override function for the next few minutes. This gives time for the AI system to check and share the lesson with the fleet. In override mode the remote pilot takes over opening the bow doors of the craft and begins the sequence that will unload the remotely operated vehicles it is carrying. 

As soon as the ramp descends a sealed container inside pops and a UAV stowed within hums into action. Immediately it zooms away from the amphibian, rising quickly to about 5 metres. It moves up the beach. Where the sand becomes soil and trees, the UAV hovers sideways. A Reccedrone7, so-called because its all up mass is below 7kg Without human direction it searches for and finds the place where a rutted track leads off the beach. There it rises higher, just clear of the canopy of mangroves and scrub. The shrouded quadcopter begins to move steadily forward directly over the track, its sensors scanning through the trees and scrub below. Other UAVs, disgorged from other submersibles, head out from the beach on separated curving tracks that keep them below the tops of the tallest trees. Back on the shoreline inside the cargo deck of the submersible, staring electronic eyes pierce both darkness and smoke, to send a view of the beach back to Australia and the operators. 

A monochrome thermal picture appears on the distant screen of the operator of the first UGV, waiting to descend from the top of the ramp. It is staccato, and endless series of static snapshots transmitted at intervals to save system bandwidth when not under human control. He sees a green light illuminate. This tells him that the pyrotechnic strips have inflated and burst the heavy polyethylene waterproof envelope protecting the amphibious platform and that the thermic shackles holding it tight onto the deck have been fired. The release allows his UGV to roll down the amphibian’s ramp and to start moving up the beach. The remote operator has rehearsed this moment forty, no maybe fifty times on the simulator, each time with a slightly different scenario, sometimes requiring intervention, sometimes leaving it to the AI. The view as the UGV heads to the edge of beach is weirdly completely familiar. The spot amongst the trees and bushes lining the beach where the little vehicle heads for is no longer a simulation. The tension he feels is shared by many of the other operators seated in clusters of consoles throughout the control room. So much depends on remaining undetected whilst the force establishes a foothold. A random Echthros patrol could change everything – and what if a civilian is sleeping undetected somewhere a bit further along the beach? 

A menagerie of airborne and ground platforms disgorge from the submersibles, varied types emerging from each one – a feature of careful cross loading that ensures a balanced capability mix of platform survives even if delivery craft are lost. 

Craft 

First the UAV’s on the ramps of the submersibles rise to disappear through the glycol smoke haze still billowing from vents on the submersibles. Then into the smoke vapour and darkness, the shadowy diffused shapes of agile ground systems follow on. 

The golf cart sized Polywheeler350 bounce and scurry up the beach on their six skid-steer wheels like frisky nocturnal animals heading for the protection of the tree line. There they pause. Were they visible to the naked eye, an observer might even suppose they were animals not machines. The UGVs’ bodies and wheels alike are irregularly covered in foam shapes and fabric strips. These are stuck to every part where they will not interfere the parts that propel sense or arm the vehicles. While moving, they appear something like an aggregation of rolling tumbleweed’s and a large seaweed encrusted turtle. When the vehicles stop, they appear to become a pile of non-descript foliage. 

Behind the agile wheeled UGVs, larger dark shapes the size of compact cars move down the submersibles’ ramps. The way they initially sink into the sandy beach, and their subsequent ponderous motion, reveals their tracks are dragging much greater weight across the sand. They too are covered with camouflage foam shapes making them look like shaggy beasts, but no animal this size has lived 12 

in these parts for millennia and the whine of electric motors is unquestionably mechanical. They are MAVOC4000’s, a name reflecting their medium weight of four tons, their armour and that they optionally crewed. Some have turrets, some have multi-barrelled launchers, but none have humans aboard. 

Several of these MAVOC move directly off the beach into the forest, crashing and smashing until they halt 50 m in. Others head for the track and with the unhesitating purpose of AI rumble up the road after the smaller UGVs. Some, mounting rocket launchers or pods to carry and recharge drones will soon deviate on side tracks towards deserted forest clearings that, like the beach landing site, have been watched by sensors for weeks. Several descend their amphibians’ ramps and turn to proceed along the artificial beach in opposite directions, going as far as they can before their path is blocked by mangrove. There they halt and festooned plastic camouflage poles swing out away from their hulls to make them more bush-like. Turrets turn to elevate twin cannons to cover the sky above the jungle canopy. As power is supplied to their airspace sensors, they begin to hum – nonhuman sentries against nonhuman flying threats. 

Even as the counter drone CUAS-MAVOC are making their way along the beach to take up protective positions, the smallest UGV amongst the first wave come ashore. They are hidden by the darkness but were they not; they are further disguised by festooned camouflage material. These vehicles echo the familiar form of the German World War II Goliath remote control mine. Like that vehicle, these Xtrak800 are conceived as expendable, but rather than being packed with explosives, these UGV are multipurpose. On their modular deck, some mount satellite dishes that provide durable data links and powerful rebroadcast for the mesh radio network, some carry cassettes to deliver and recharge smaller UAS and UGV, and others mount an elevating multi-tube weapon array. Unlike the Goliath locomotive design, these have wider rubber tracks that ensure they can cross most uneven ground. Each is connected to a wheeled supplementary-battery trailer on a spring-loaded arm that improves gap and step crossing capabilities but can be disconnected when no longer required. Their hulls, running assemblies and system modules are fabricated from tough steel plate, armouring them against small-arms fire. The resultant mass of nearly a ton is obvious in their slower speed and ponderous motion as they crawl down the ramp of the Sublanders and advance across the beach to the track. Most of the Xtrak advance steadily until they close up behind the leading Polywheeler350s. 

The last few vehicles to come ashore are different, their four-wheel chassis’ piled with fabric and panels. Halfway up the beach the first one stops and begins to swell, wobbling before the compressed air coursing around its skeleton of inflation tubes suddenly pops it up into the form of a four-wheel-drive. To the casual glance it looks like a battered Toyota SUV, but the easy way it begins to move across the sand again reveals its true nature: an inflated folding plastic shell on a lightweight electrically powered UGV. It moves up the beach to where the UAV disappeared down the track and follows it. As it departs, the fake car is followed by several of the Polywheeler350, the one immediately behind the Toyota blows glycol smoke in the darkness. With the naked human eye, an observer would see little through the rolling fog, but if they did see anything, it would appear to be an unremarkable civilian vehicle. 

Kilometres away, further back along the murky underwater course the pod of Sublanders took, two of their kind have deviated. They have dragged themselves, sloshing through the mud and mangrove roots, to lay half out of the water. No vehicles appear on their ramps. Instead, there is a steady succession of dull popping noises announcing the launch of can-sized projectiles into the sky, unseen in the darkness: Dropdrone4’s munition delivery UAS. As each ejects from its launch tube, spinning motors drive propeller blades to unfold and claw at the warm night air. The devices hang for a moment to stabilise and to get their bearings from distant satellite clocks, before humming off above and across the mangroves and scrub. Each one has a full set of waypoints, destinations and observation tasks stored in its electronic brain – and a provisional selection. These self-placing sensor carriers, without reference to human control, operate as a ‘linear swarm’. In their turn, they whine off on a staccato course across the treetops, 

The first Dropdrone4 arrives

over its destination in an overgrown dried paddy, thick with rice stalks. Suddenly, the bladed transport cylinder shoots upward, freed from the weight of its payload as the sensor assembly it carried drops away and plunges to embed itself in the earth. A few seconds pass as the device uses the presence of the hovering carrier above it to calibrate its microphones, then it sends a single short message to the swarm and onward to Home Control: in position. Absent that success transmission, another device further back in the queue would have reassigned itself to that spot without transmission from Darwin. Most of the sensors are placed in the ground and there concealed by foliage, but a few devices are placed on top of buildings. Just above the roof the carrier drops a device sprouting feathers like a shuttlecock which ensures that the viscous adhesive pad on the under-surface strikes and sticks. Concealed beneath a mass of more fake feathers that make the device look somewhat like a dead bird, is a camera, processor and transmitter. 

Back on the beach, as each of the Sublanders discharges its last platforms, the distant pilots initiate a departure sequence which raises the bow door, switches off the smoke generator and begins to reverse the amphibian back to sea. Already, the navigation systems are communicating with each other, so that when the tracks are clear from the seabed, and the platforms begin to wallow in the waves the AI will keep them clear of each other: machine computational thinking is far better at coordinating a dozen submerged platforms reversing in a small fluid space than humans are. Within minutes the collective intelligence has the submersibles back just below the water’s surface and heading out to sea, aided by the river current. On the way the collective intelligence of the pod keeps the swarm of unladen submersibles well clear of where a second wave of craft lie static on the riverbed. 

 

Chapter 2 is here

Thanks to Electro Optic Systems for helping create this content.

Original Artwork imagined by Charles Knight, drawn by James Knight