Construction machines: longer booms, lighter steel, few operators

Longer booms on concrete pumps and platforms vibrate

What happens. Concrete pump booms and aerial platforms keep getting longer, and made of lighter steel. The boom swings when it moves or when the pump pulses. The operator sees the tip tremble and slows down.

What it costs. Time at every placement, a safety risk when the tip swings near people, and a boom that cannot use its full reach at full speed.

Why the usual fix fails. Stiffening adds weight and cost. Operator skill in moving smoothly cannot be built into a machine. A simple damper tuned for one boom position does not hold at the next.

How a fixed-price remote diagnosis checks it. From sensor and valve logs of boom movements I identify the swing modes and how they change with position. The fixed-price remote diagnosis tells you whether active damping can be added with the sensors you have.

Longer booms and lighter steel let the load sway

What happens. On cranes with a long, light boom the load hangs on a rope and swings after each movement. The operator waits for it to settle or counters it by feel.

What it costs. Cycle time, risk in tight spaces, and wear on the slewing gear from hard braking.

Why the usual fix fails. Slower movements are safe and unproductive. A sensor for the rope angle shows the sway but does not tell the drive what to do. Fixed filters in the operator’s command work only for one rope length.

How a fixed-price remote diagnosis checks it. From crane logs (drive commands, angles, rope length, load) I look at how the sway develops. The fixed-price remote diagnosis tells you whether estimation of the sway with predictive control of the drives can cut cycle time.

The load is not measured precisely, so safety margins are wide

What happens. The load on a crane or a lifting machine is judged from hydraulic pressures and geometry. Friction, temperature and boom position make it inexact. Load-moment limits are therefore set with margin.

What it costs. Lifting capacity that is not used, early cut-outs that annoy operators, and sometimes a false sense of safety.

Why the usual fix fails. Load cells add cost and wiring in a rough environment. Wider margins give capacity away. Calibrating the pressure signal once does not follow oil temperature and wear.

How a fixed-price remote diagnosis checks it. From pressure, position and test lift data I look at how well the load can be computed and how accurate it is. The fixed-price remote diagnosis tells you whether the margins could shrink with an estimated load.

Background: Computed measurements

Excavator grading depends on the operator's skill

What happens. To dig to a level, a slope or a trench depth, the operator moves boom, stick and bucket together. A good operator does this smoothly. A new one overshoots and corrects.

What it costs. Rework and extra material, slow work in finishing, and a dependence on the few good operators.

Why the usual fix fails. Depth displays show the error but leave the movement to the operator. Training takes years. Assistance with fixed rules works on flat, easy ground and not on a real site.

How a fixed-price remote diagnosis checks it. From machine logs of digging cycles (joint angles, cylinder pressures, bucket position) I look at how much automated grading could follow the target. The fixed-price remote diagnosis tells you what sensing and control the function needs.

Background: Robotic manipulation with precise predictive control

Hydraulic pumps run at full pressure

What happens. The pump delivers oil at the highest pressure any function might need, and valves throttle it down for lighter jobs. The difference becomes heat.

What it costs. Fuel, oil heating that needs cooling, and wear in throttling valves.

Why the usual fix fails. Load-sensing helps but still runs with a pressure margin. A variable pump with a fixed setting is an improvement and not an optimum. Rewriting the whole hydraulics is out of reach for a running series.

How a fixed-price remote diagnosis checks it. From duty-cycle logs (pump pressure, flow, engine speed, function pressures) I look at how much energy goes into throttling. The fixed-price remote diagnosis tells you whether an optimised load-sensing control would save fuel.

Background: Economic predictive control

Battery excavators run out of range

What happens. A battery excavator carries a limited amount of energy. A heavy digging shift can drain it before the day is over, and the operator drives without knowing what is left for the job.

What it costs. Downtime for charging, a diesel machine kept as back-up, and customers who size the battery for the worst day.

Why the usual fix fails. A bigger battery adds cost and weight. An energy display shows the past, not the plan. Eco modes with fixed power limits slow every job.

How a fixed-price remote diagnosis checks it. From operating logs (battery state, power per function, job type) I look at how well the energy for a task can be predicted. The fixed-price remote diagnosis tells you whether energy management that plans work and charging can extend usable shifts.

Background: Economic predictive control

Engine and battery in hybrid drives are not coordinated

What happens. A hybrid machine has an engine, a battery and an electric motor, but the split is made by simple rules. The engine runs in poor points while the battery sits idle, or the other way round.

What it costs. Fuel that a better split would have saved, and a hybrid benefit that does not reach the figure promised.

Why the usual fix fails. Rules for each duty cycle must be written again for each machine. Tuning on the test cycle does not carry over to real work. Forecasting the load is not part of the control.

How a fixed-price remote diagnosis checks it. From duty-cycle logs and the drive data I look at what the best split would have saved over the real work. The fixed-price remote diagnosis tells you whether predictive energy management is worth building.

Background: Economic predictive control

Tower crane work stops in wind

What happens. Tower cranes stop lifting when the wind passes a limit. The limit is set for the worst case of load and wind direction, and gusts make operators stop even when the next minutes would have been fine.

What it costs. Idle site days, delayed schedules and crews waiting.

Why the usual fix fails. A conservative fixed limit protects and costs hours. Raising it by decision of the site manager risks an accident. A wind sensor at the top shows the wind there, not what hits the load.

How a fixed-price remote diagnosis checks it. From wind records, crane data and load types I look at how often the limit is decisive and what a load- and direction-aware limit would release. The fixed-price remote diagnosis tells you whether wind-aware control can allow more lifting hours safely.

Tipping margins on slopes are large

What happens. A machine working on a slope with a raised boom or a full bucket can tip. Makers set the margins large, so the machine is held back long before it tips.

What it costs. Productivity lost on slopes, an operator who sees the machine refuse a movement without reason, and sometimes still an accident.

Why the usual fix fails. A slope sensor gives the angle and nothing about the dynamic effect of moving the load. Larger counterweights cost money. Fixed limits for the worst case apply on every job.

How a fixed-price remote diagnosis checks it. From machine data (inclination, joint angles, cylinder pressures) I look at how well the real stability margin can be computed during work. The fixed-price remote diagnosis tells you whether the limits can follow the computed margin.

Background: Computing unmeasured states for control

Drilling parameters are set by feel

What happens. A drilling rig is run with feed force, rotation and flushing set by the operator’s experience. The rock changes along the hole, and the settings follow slowly.

What it costs. Bit wear, slow drilling speed, stuck tools and a different result from shift to shift.

Why the usual fix fails. Fixed parameter sets suit one rock type. An experienced drill operator cannot be everywhere. Sensors show forces but not what to change.

How a fixed-price remote diagnosis checks it. From drilling logs (feed, torque, rotation, penetration rate, wear records) I look at how the rock shows in the signals and how the parameters relate to speed and wear. The fixed-price remote diagnosis tells you whether optimised drilling control is possible.

Background: Economic predictive control

Road pavers struggle with mat thickness and smoothness

What happens. A paver lays an asphalt mat of defined thickness. Material temperature, speed and screed movement affect the result. The screed floats on the material and reacts slowly to changes.

What it costs. Rework, extra material, and surfaces that are not smooth enough, which can mean deductions for the contractor.

Why the usual fix fails. Manual screed corrections come after the error has been laid. Fixed settings suit one mix and speed. Sensors on the screed show the position but not the cause.

How a fixed-price remote diagnosis checks it. From paver logs (screed position, angle of attack, speed, material and thickness measurements) I look at how the screed responds to changes. The fixed-price remote diagnosis tells you whether model-based screed control can improve thickness and smoothness.

Background: Industrial process predictive control

Compaction is measured afterwards

What happens. A compactor rolls over the mat several times. How well it is compacted is measured afterwards, with a probe or a test. In between, the operator decides by experience when to stop.

What it costs. Over-compaction that wastes passes and damages the material, under-compaction that shortens the life of the road, and rework.

Why the usual fix fails. Measuring afterwards finds errors when the roller has gone. Fixed numbers of passes ignore material and temperature. Roller sensors give a stiffness signal that is hard to read.

How a fixed-price remote diagnosis checks it. From roller data (acceleration, speed, position) and test results I look at how well compaction can be estimated during rolling. The fixed-price remote diagnosis tells you whether a compaction estimate can tell the operator when to stop.

Background: Computed measurements

Remote operators work with lag and lose precision

What happens. In teleoperation the operator sits elsewhere and sees the machine through a camera. Video and commands arrive with delay, and precise work becomes hard.

What it costs. Slow work, mistakes in fine moves, and a dependency on the quality of the connection.

Why the usual fix fails. Faster connections shorten the delay and do not remove it. Operators slow down to compensate. Pure camera views lack the feel of the machine.

How a fixed-price remote diagnosis checks it. From recordings of remote operation with delays I look at which moves suffer most. The fixed-price remote diagnosis tells you whether predictive assistance that bridges the delay can help.

Background: Predictive control with obstacle avoidance under uncertainty

Skilled machine operators are scarce

What happens. Contractors cannot find experienced operators. New ones work slower, use more fuel and wear the machine harder.

What it costs. Lost productivity, more damage to machines, and customers who ask for machines that “run themselves”.

Why the usual fix fails. Training takes years. Higher pay alone does not produce people. Complex machines raise the level needed.

How a fixed-price remote diagnosis checks it. From machine logs of experienced and newer operators I look at what the best operators do differently and which of it can be taken over by the control. The fixed-price remote diagnosis tells you which assistance functions have the largest effect.

A bigger machine arrives with the old control

What happens. The next machine generation is larger and heavier, but the control comes from the smaller one. The hydraulics, the boom and the inertia differ, and the old gains no longer fit.

What it costs. Instability and oscillations in trials, repeated tuning on the prototype and a later release.

Why the usual fix fails. Scaling gains by size is a guess. Tuning every function on the new machine is slow. The experts who knew the old tuning have moved on.

How a fixed-price remote diagnosis checks it. From the old machine’s data and the new machine’s design values I look at which dynamics change with size. The fixed-price remote diagnosis tells you what a control for the new machine needs from the start.

Background: Calibration-efficient predictive control

Structures crack early under unknown loads

What happens. Booms and frames crack earlier than calculated. The loads in real work are higher than assumed in the design, because operators hit stops, drop loads and work at the limit.

What it costs. Warranty claims, repairs on site and damage to the brand’s reputation.

Why the usual fix fails. Stronger steel adds weight and cost. Strain gauges on a few machines give a view of a few jobs. Warranty cases come years after design.

How a fixed-price remote diagnosis checks it. From pressure and motion data across machines I look at whether the loads in the structure can be estimated, and which operation causes the peaks. The fixed-price remote diagnosis tells you whether load-limiting control can protect the structure.

Background: Computing unmeasured states for control

Mobile controllers are too limited for better algorithms

What happens. Controllers in mobile machines have little memory and computing power and strict requirements for temperature and vibration. The algorithms that would help are too heavy for them.

What it costs. Simple algorithms that leave performance unused, and a gap between what is researched and what is built into the series.

Why the usual fix fails. A more powerful controller costs more and requires a new supply chain. Rewriting by hand for the limited processor is slow. Cutting the algorithm without checking loses its benefit.

How a fixed-price remote diagnosis checks it. From the function description and the processor limits I look at how much of the algorithm can run in the time budget and what it loses by simplifying. The fixed-price remote diagnosis tells you whether real-time code on your controller is feasible.

Background: Predictive control on embedded hardware

Field tests are expensive and slow development

What happens. Many functions can only be judged on a real machine on a real site. Tests need ground, crew, fuel and time, and they depend on weather.

What it costs. Development time and cost, and release dates that follow the test calendar.

Why the usual fix fails. More test machines multiply the cost. Testing only on the test track misses real ground. A simulation without matching to field data is not trusted.

How a fixed-price remote diagnosis checks it. From field logs and your existing models I look at how well a simulation reproduces the machine and which tests it could replace. The fixed-price remote diagnosis tells you what simulation-first development would need.

Breakdowns on site stop the job

What happens. A machine breaks down on site. The fault code says little, the service technician travels without knowing the cause, and the job waits.

What it costs. Downtime for the customer, a second service visit for a missing part, and an unhappy reputation.

Why the usual fix fails. Fault codes show what tripped, not why. Fixed service intervals ignore real use. Remote data collection without analysis only fills a database.

How a fixed-price remote diagnosis checks it. From fault logs and measured values around past breakdowns I look at whether failures show early signs. The fixed-price remote diagnosis tells you whether on-board diagnosis can warn before the stop and tell the technician what to bring.

Mobile machines need a control concept from idea to series

What happens. Makers of mobile machines such as chippers and shredders build machines in which engine, hydraulics and automation must work together. A small team has to take the control from the first concept to a series product.

What it costs. A concept that grows without a plan, late changes and a start of production that depends on a few people.

Why the usual fix fails. Adding control functions one by one yields a tangle. Buying a standard controller does not give the machine-specific logic. Hiring the right person takes time the project does not have.

How a fixed-price remote diagnosis checks it. From the machine description, the process and existing control code I look at which functions are needed and how they can be structured. The fixed-price remote diagnosis tells you what a clean control concept looks like and what remains to be built.

Electric construction sites are limited by site power, not by the machines

What happens. Electric construction machines exist, but on site the bottleneck is power: the grid connection, the charging time and the energy available. Battery size is then weighed against shift length and hydraulic losses.

What it costs. Machines waiting for a charge, oversized batteries and connections, and contractors who stay with diesel.

Why the usual fix fails. A bigger grid connection is slow and expensive. Charging on the machine’s own schedule piles up peaks. Fixed charging times ignore what the day’s work needs.

How a fixed-price remote diagnosis checks it. From machine usage logs and site power data I look at the load profile of the machines and the connection. The fixed-price remote diagnosis tells you whether energy management of the battery machines and a charging plan for the site can cut the needed battery and connection size.

Background: Economic predictive control