Commercial robots are moving out of highly controlled manufacturing environments and into the ordinary places where people work.
They transport materials through warehouses and factories. They move food through restaurants. They clean floors in offices, hotels, hospitals and public buildings. They deliver supplies. They provide information to customers. They perform repetitive tasks that, until recently, could only realistically be performed by people.
This inevitably raises a much larger question.
What happens to human work when machines become capable of doing more of it?
At one extreme are predictions of a future in which robots and artificial intelligence perform virtually all economically productive work and employment becomes largely unnecessary. At the other is an equally questionable assumption that technology changes occupations without meaningfully displacing workers.
Neither extreme provides much help to a business deciding whether robotic automation makes sense today.
The practical question is more immediate:
Where can robots perform repetitive, predictable or labor-intensive work more economically, reliably or safely than current methods—and how should businesses combine that capability with the things people continue to do better?
That is the economic case for commercial robotics worth examining.
It begins not with the robot, but with the work.
Robots themselves are not new.
Manufacturers have used industrial robots for decades for welding, assembly, painting, material handling and other repetitive processes. What is changing is the range of environments in which robotic automation has become commercially practical.
Advances in sensors, cameras, LiDAR, computer vision, navigation, batteries, wireless connectivity, cloud computing and artificial intelligence have enabled a new generation of machines that can operate in environments designed primarily for people.
At the same time, the commercial model is changing.
Businesses increasingly have access to robots through financing, leasing and Robot-as-a-Service arrangements rather than having to treat automation as a large engineering project requiring substantial upfront capital.
This combination is significant.
According to the International Federation of Robotics, worldwide sales of professional service robots approached 200,000 units in 2024. Transportation and logistics accounted for more than half of those units. More than 42,000 hospitality robots were sold, while professional cleaning robot sales exceeded 25,000 units and increased 34% from the previous year.
Robot-as-a-Service fleets are also expanding, lowering the capital barrier to adoption.
These figures should be interpreted carefully—the IFR service-robot data represents participating suppliers rather than a projection of the entire global market—but the direction is clear.
Commercial robotics is no longer simply an experimental technology looking for applications.
Businesses are finding applications where the economics already work.
Discussions about robotics often begin with occupations.
Will robots replace cleaners?
Will they replace waiters?
Will they replace warehouse workers?
That framing can be misleading because businesses generally do not automate an occupation all at once.
They automate tasks and workflows.
Consider commercial cleaning.
A large facility may require employees to spend many hours repeatedly traversing corridors and open floor areas with cleaning equipment. An autonomous cleaning machine may be able to perform much of that repetitive floor coverage.
But the facility still needs bathrooms cleaned, waste removed, supplies replenished, spills handled, corners and inaccessible areas addressed, equipment maintained and unexpected situations resolved.
The robot has not necessarily replaced “the cleaner.”
It has automated a portion of the cleaning workload.
The same distinction appears across many emerging commercial applications.
Autonomous machines can transport meals, dishes, linens, supplies or other items between defined locations.
The economic opportunity is not necessarily eliminating the employee who previously moved those items.
It may be reducing the amount of employee time consumed simply walking products from one place to another.
Robotic machines can sweep, vacuum, scrub or maintain large floor areas according to scheduled routes.
Employees can concentrate on tasks that remain difficult to automate or require judgment and manual dexterity.
Autonomous mobile robots can transport components, tools, inventory and materials between workstations, storage locations and production areas.
The objective may be to reduce repetitive transportation rather than automate the production job itself.
Robots can provide information, wayfinding, promotional content or basic interaction in retail, hospitality and public environments.
Again, these functions may supplement rather than eliminate human customer service.
This distinction matters enormously when evaluating the economics.
The useful unit of analysis is not the job title. It is the task.
What work is being performed?
How many hours does it consume?
What does that work cost?
How predictable is it?
How much human judgment does it require?
And what would happen to the rest of the employee's workload if that particular task were automated?
Those questions produce a far more useful automation strategy than simply asking how many employees a robot can replace.
The simplest robotics sales pitch is also one of the most dangerous:
A robot costs less than an employee.
Sometimes it does.
But that comparison is too simplistic to support a serious investment decision.
A business should compare the total cost of performing the work today with the total cost of performing that work after automation.
An employee's economic cost is more than the hourly wage.
Depending upon the employer and position, labor costs can include:
wages or salary;
employer payroll taxes;
benefits;
workers' compensation;
overtime;
recruitment;
onboarding and training;
uniforms or equipment;
supervision;
absenteeism;
turnover and replacement;
temporary staffing;
scheduling inefficiencies.
The wage itself nevertheless provides useful context.
Current U.S. Bureau of Labor Statistics data puts mean hourly wages at approximately $18–$20 for several occupational groups directly relevant to emerging commercial robotics, including building cleaning workers, food-service workers and material-moving workers.
The fully burdened employer cost can be materially higher.
A robot isn't free labor either.
Its economics can include:
purchase price or lease payment;
financing;
installation;
site assessment;
mapping and configuration;
workflow integration;
employee training;
software or subscription charges;
connectivity;
electricity;
consumables;
preventive maintenance;
repairs;
replacement parts;
downtime;
eventual replacement.
That means the correct comparison is not:
Hourly wage vs. robot payment.
It is:
Current total cost of accomplishing the work vs. future total cost of accomplishing the same or greater amount of work with automation.
That calculation also needs to account for productivity.
If a robot costs $800 per month but performs only $400 worth of useful work, it is not economical merely because $800 is less than a full-time employee's salary.
Conversely, if the same machine consistently performs 150 hours of repetitive work each month, the economics can become compelling even when no employee is actually eliminated.
The organization may instead use those recovered hours elsewhere.
Suppose a restaurant employee spends part of every shift moving food from the kitchen to tables.
A delivery robot may be able to perform much of that movement.
Does the restaurant therefore need one fewer employee?
Possibly.
But that is only one potential economic outcome.
The restaurant might instead use the recovered employee time for greeting customers, taking orders, clearing tables, selling additional drinks, handling exceptions or improving service.
The economic return from the robot could therefore come from:
labor reduction, labor avoidance, increased capacity, better service—or some combination of all four.
The same applies to commercial cleaning.
An autonomous floor-cleaning machine may reduce the number of labor hours required to maintain a building. An employer facing persistent vacancies may use the robot to fill that capacity gap rather than reduce headcount.
Another employer may reduce overtime.
Another may redeploy employees to higher-value cleaning tasks.
Another may use automation to increase the frequency with which the facility is cleaned.
Another may genuinely require fewer employees.
All are forms of economic impact.
This is why the question:
“How many employees does this robot replace?”
should usually be replaced with:
“How does this robot change the economics of this workflow?”
That is a much more useful business question—and a much more honest one.
Some of the strongest automation opportunities occur when robots increase the productive capacity of existing employees.
Imagine a facility-cleaning team responsible for a large building.
If autonomous equipment handles predictable open-floor cleaning, employees can perform work requiring greater dexterity, judgment or interaction while the robot operates simultaneously.
The organization has effectively added productive capacity without adding another conventional labor shift.
A similar effect occurs in material movement.
If a skilled production employee repeatedly leaves a workstation to retrieve components, the cost of that movement isn't simply the employee's walking time.
The organization is also losing productive capacity from the activity that employee was hired to perform.
Automating transportation can therefore generate value even when total employment remains unchanged.
This leads to an important principle:
The value of automation is not limited to the labor it replaces. It includes the productive human capacity it releases.
That capacity can produce several outcomes:
greater throughput;
increased facility coverage;
longer operating hours;
more consistent service;
reduced overtime;
shorter customer wait times;
increased revenue capacity;
reduced physical workload;
improved use of skilled employees.
This is the concept of human augmentation rather than simple human substitution.
The distinction will become increasingly important as robotics, AI and connected systems become more deeply embedded in ordinary business operations.
Sometimes the economics of automation are driven less by the cost of employees than by their availability.
Businesses cannot deploy labor they cannot hire.
The International Federation of Robotics specifically identifies staff shortages as one of the drivers behind professional service-robot adoption.
Demographics may reinforce the issue.
The U.S. Bureau of Labor Statistics projects the civilian labor force to grow only 3.2% between 2024 and 2034, while the overall labor-force participation rate is projected to decline from 62.6% to 61.1%.
That does not mean every industry faces a labor shortage or that robots provide the answer wherever shortages exist.
But it changes the automation calculation.
An employer may face:
chronically unfilled positions;
high employee turnover;
expensive overtime;
dependence on temporary labor;
difficulty staffing nights or weekends;
inconsistent service caused by staffing gaps;
repetitive or physically demanding jobs that are difficult to retain.
In those circumstances, the relevant question may not be:
“Should we replace this employee with a robot?”
It may be:
“How do we reliably accomplish work for which we cannot maintain sufficient labor capacity?”
Robotics can become an additional source of operating capacity.
And unlike the theoretical debate about a future without work, that is a problem many businesses face today.
This is the question the robotics industry should not avoid.
Automation can displace workers.
There is substantial economic research demonstrating that effect.
Research by economists Daron Acemoglu and Pascual Restrepo examining U.S. labor markets found negative employment and wage effects associated with increased exposure to industrial robots. Other firm-level research has found that companies adopting robots can become more productive and expand while competitive effects elsewhere in the industry can still produce an overall reduction in employment.
That matters.
It would be misleading to argue that every worker displaced from an automated task simply moves into a better job.
Economic transitions are rarely that frictionless.
People can lose jobs. Skills can become less valuable. Communities and occupations can experience disruption. Workers may need retraining. The benefits and costs of technological change are not necessarily distributed evenly.
But that evidence does not support the opposite extreme either: that robots inevitably lead to the disappearance of human employment.
The economic effects work in several directions simultaneously.
Automation can displace labor by allowing machines to perform tasks previously performed by people.
It can also produce a productivity effect. Lower costs or greater productive capacity can increase output and create demand for work elsewhere.
Technology can also create new tasks and occupations that previously did not exist.
The OECD describes essentially these competing displacement, productivity and new-task effects and notes that their ultimate impact on total labor demand cannot simply be assumed in advance.
The World Economic Forum's 2025 employer survey illustrates the same complexity. It projects considerable labor-market disruption through 2030, including both substantial job creation and substantial displacement. Robotics and autonomous systems are expected to be net job displacers within that analysis—but against a wider backdrop in which employers anticipate overall net job creation from the major forces reshaping the economy.
The sensible conclusion is therefore neither:
“Robots won't cost jobs.”
nor:
“Robots will eliminate the need for human work.”
It is:
Robotics will change the demand for human labor, and the consequences will differ substantially by task, occupation, industry and business.
Predictions of a future in which almost nobody needs to work also raise an obvious economic problem.
Businesses automate partly to produce goods and services more efficiently.
But businesses ultimately require customers.
Customers require purchasing power.
If technological automation were eventually to remove the overwhelming majority of earned income from an economy, society would need different mechanisms for distributing purchasing power or the economic system itself would have to change substantially.
That is an important debate.
It is not, however, a question a restaurant, warehouse, hotel or commercial facility must solve before deciding whether to automate repetitive work in 2026.
A business investing in a commercial robot today is generally making a decision over a foreseeable investment horizon.
It needs to understand what the machine costs, what work it can perform, what operating savings or capacity it can produce, how reliable it is, how it will affect employees and how quickly the investment pays back.
It does not need a theory of the post-work economy to make a rational automation decision.
The more plausible near-term workplace is not one without people.
It is one in which people work alongside an expanding collection of intelligent machines and software.
That environment may contain:
employees + autonomous robots + AI systems + remote specialists + connected field technicians + predictive analytics + automated monitoring.
This changes more than staffing levels.
It changes how work itself can be organized.
Consider a commercial robot operating in a customer's facility.
The robot performs its assigned work.
Operational data can potentially identify degraded performance or an emerging problem.
Remote diagnostics may resolve a software or configuration issue without a site visit.
If physical service is required, a technician can arrive with information about the likely failure rather than beginning the diagnostic process from scratch.
If the technician encounters an unfamiliar or complex issue, augmented-reality tools can connect that technician with a specialist elsewhere.
The specialist can see what the field technician sees and provide real-time guidance.
The result can be:
fewer unnecessary service calls;
better preparation before dispatch;
higher first-visit fix rates;
shorter repair times;
more efficient use of specialist technicians;
greater equipment uptime.
This is an important reminder that technological change does not always remove human capability from a process.
Sometimes it amplifies it.
A technician supported by AI, remote diagnostics, operational data and augmented reality may be substantially more productive than the same technician working alone.
The economic question therefore extends beyond what robots can automate.
It includes what technology allows people to do better.
A robot generates economic value only while it is performing useful work.
That makes reliability and service part of the ROI calculation.
Suppose an automation project is expected to recover 120 labor hours each month.
If the robot is unavailable for a significant portion of that time, the projected return deteriorates rapidly.
Worse, the business may still need to maintain enough human capacity to perform the work whenever the automated system is unavailable.
That makes several questions financially important before deployment:
Who supports the machine?
How quickly can a problem be diagnosed?
Can issues be resolved remotely?
Are replacement parts available?
Is local field service available?
What preventive maintenance is required?
What is the expected response time?
How frequently will the machine be unavailable?
How quickly can it return to productive service?
These are not merely service questions.
They are economic variables.
Two robots with similar purchase prices and specifications can produce very different returns if one operates reliably with effective local support while the other experiences prolonged downtime.
The probability of equipment failure can never be reduced to zero.
A more practical objective is to minimize both:
how frequently failures occur and how long they interrupt productive operation.
Remote monitoring and predictive analytics can help identify problems earlier.
Remote support can resolve some failures without dispatch.
Better diagnostics can ensure technicians arrive with appropriate knowledge and parts.
Augmented-reality support can connect a field technician to deeper expertise without requiring a second specialist visit.
All of these capabilities influence the real economics of automation.
This leads to one of the most important principles in evaluating commercial robotics:
Do not evaluate the robot independently from the service infrastructure required to keep it productive.
The purchase price is visible.
The cost of downtime often isn't—until the machine stops.
Robotics ROI should begin with the existing workflow.
Not with the robot.
Determine:
what task is being performed;
how many employees participate;
how many labor hours the task consumes;
how frequently it occurs;
the fully burdened hourly labor cost;
overtime attributable to the task;
temporary staffing costs;
turnover and recruitment costs where relevant;
current throughput or output;
operational problems caused by inadequate staffing.
This establishes the current cost of the work.
Few commercial robots automate 100% of a complex workflow.
Estimate realistically:
percentage of the task the robot can perform;
expected productive operating hours;
human intervention required;
work that remains manual;
additional employee capacity created;
any increase in throughput or service levels.
This is where overly optimistic ROI projections often begin.
A robot capable of technically operating eight hours per day is not necessarily producing eight hours of economically useful work.
Include:
equipment;
financing or leasing;
deployment;
integration;
training;
software;
connectivity;
consumables;
maintenance;
expected repairs;
service agreements;
energy;
other recurring costs.
Potential benefits may include:
Actual reduction in labor hours or positions.
Future hiring that is no longer required as the organization grows.
Work transferred to automated capacity.
Existing employees redirected to higher-value work.
More deliveries, greater floor coverage, increased throughput or longer operating hours.
Tasks completed reliably according to schedule.
Physically repetitive movement transferred from employees to machines.
Not every benefit should automatically be converted into dollars. ROI models become less credible when every possible advantage is assigned an optimistic financial value.
Start with benefits that can actually be measured.
Useful measures include:
Monthly operating savings
Annual operating savings
Cost per productive hour
Payback period
Break-even point
Three-year ROI
Five-year ROI
A useful robotics ROI model should also allow assumptions to change.
What happens if the robot performs only 70% of the expected work?
What happens if wages rise?
What happens if service costs are higher than anticipated?
What happens if utilization increases?
What happens if the business grows?
That sensitivity analysis is often more valuable than a single headline ROI percentage.
The purpose of the calculation isn't to prove that robotics works.
It is to determine whether robotics works for this particular operation.
Not every task should be automated.
That may sound obvious, but enthusiasm for new technology can encourage businesses to begin with the solution and search afterward for a problem it can solve.
The better process works in the opposite direction.
Begin with an expensive, repetitive or operationally difficult problem.
Then determine whether automation improves it.
Robotics may be a poor fit where:
task volume is too low;
the environment changes constantly;
workflows are highly unpredictable;
substantial human judgment is required;
customer interaction is central to the task;
physical conditions make autonomous navigation difficult;
human intervention is required too frequently;
integration costs are disproportionate to potential savings;
the machine would have low utilization;
service support is inadequate;
expected payback exceeds a reasonable investment horizon.
There are also situations where a technically feasible deployment makes little operational sense.
A robot may be capable of transporting materials across a facility, for example, but if an employee performs the same movement only several times per day and each trip takes two minutes, the economic opportunity may be negligible.
Conversely, the same machine operating continuously across multiple shifts could have an entirely different business case.
This is why utilization matters so much.
Buying an automation asset that spends most of its life idle is rarely an attractive investment.
A credible automation assessment should be capable of producing three answers:
Yes.
Not yet.
No.
If the only possible outcome of an automation consultation is a recommendation to buy a robot, the assessment isn't really an assessment.
The objective is not robotic adoption.
The objective is better business performance.
Robotics is one tool for achieving it.
The best place to begin is not with a robot catalog.
Walk through the business.
Look for work.
Specifically, look for work that is:
repetitive, measurable, predictable, labor-intensive, difficult to staff or operationally necessary.
Then ask:
Measure them rather than estimating from memory.
Persistent vacancies can make automation economically attractive even without reducing current headcount.
A task that continually requires recruitment and training has costs beyond wages.
Moving food, inventory, supplies, parts or equipment repeatedly between known locations can be particularly suitable for autonomous systems.
Some work creates value simply by happening reliably and on schedule.
A highly paid employee walking back and forth to retrieve materials may represent a stronger automation opportunity than a lower-paid employee already working efficiently.
Calculate the fully burdened operating cost.
Do not assume 100%.
Every automation plan needs a failure-state workflow.
Deployment is the beginning of the automation lifecycle, not the end.
Model conservative, expected and optimistic scenarios.
This question belongs in the business case too.
Will positions be eliminated?
Will hiring requirements decline?
Will employees be reassigned?
Will the technology reduce physical strain?
Will employees require new skills?
How will the change be communicated?
Good automation planning considers both the financial spreadsheet and the operating organization that must live with the decision.
Commercial robotics is likely to become increasingly common because the underlying economics are becoming attractive across a growing range of applications.
That does not mean every robot saves money.
It does not mean every task should be automated.
And it certainly does not require believing that human employment is approaching extinction.
The more practical future is already emerging.
People will continue performing work where judgment, dexterity, creativity, empathy, accountability and adaptability matter.
Machines will increasingly perform work that is repetitive, measurable, predictable or physically inefficient for people to perform.
Artificial intelligence will help both make better decisions.
Connected systems will monitor equipment and operations.
Remote expertise and augmented reality will make field employees more capable.
And businesses will continuously reconsider where the boundary between human work and machine work should sit.
That boundary will move.
Sometimes it will eliminate jobs. Sometimes it will eliminate vacancies. Sometimes it will remove undesirable tasks from existing jobs. Sometimes it will allow the same number of employees to accomplish considerably more.
There is no single answer because there is no single automation problem.
For a business considering commercial robotics, that broader uncertainty should not prevent a rational decision.
The questions are much more concrete:
What work are we trying to improve?
What does that work cost today?
What portion can realistically be automated?
What will automation cost?
How will our people work differently?
How will the technology be supported?
What happens when it fails?
And what return should we reasonably expect?
Answer those questions well and the decision becomes less about the future of robots.
It becomes what it should have been from the beginning:
a business decision about the future of the operation.
The right commercial robot begins with the right business problem.
Evaluate repetitive workflows, labor requirements, operating costs, utilization and service requirements before selecting a technology or manufacturer.
Calculate the potential ROI of commercial automation, explore applications by business environment, or ask the Apollo Command Center how robotics could apply to a specific workflow.