Costs of internal transport automation
The readiness of a company to make changes related to digitization in the context of logistics and Industry 4.0 means that we have the appropriate data to make informed decisions. Therefore, we will start by examining the data. A lack of proper information at the outset means:
- Inability to assess the effectiveness of human and robot work
- Certainty that available resources are not optimally utilized
- Very difficult planning and unrealistic schedules
Many companies believe that automation is the right direction to increase their efficiency. However, if they are not prepared for it, the effect can be the opposite of what was intended. How can you tell if a company is ready for new technologies? It can be indicated by:
- Consistently measured efficiency
- Standardized processes
- Constantly optimized production efficiency
- Low percentage of unplanned downtime
- Low cost of waste/materials loss
In the industry that implements robots, it is popular to say that everything can be automated, including inefficiency and generating losses. If the OEE (Overall Equipment Efficiency) of a particular workstation is 30%, its automation will not raise it to 100%. We will simply have 30% robotized, potentially much more efficient process. From our observations, the most important criteria that should determine whether it is worth analyzing the implementation of mobile robots are:
- High employee turnover and/or difficulty in recruiting them
- Rising labor costs
- Consequences of poor quality work (a large number of accidents/damage to goods)
- Overpriced internal logistics costs compared to the competition
An additional reason may be the certainty, based on data, that despite the use of best practices, the team of manual workers has reached a ceiling and further development can only be enabled by the implementation of new technologies. Automation should, therefore, be preceded or complemented by event measurement and process standardization. It is important that the data is collected in as consistent a manner as possible and through various sensors, without the need to involve employees who may skip their input or modify it according to their discretion.
It is worth remembering that the automation of internal transport will best demonstrate its capabilities when integrated with the company's tools for processing collected data and ERP/WMS systems.
The proposed analysis will examine two areas that will allow for a decision on robotization based on reliable data. We will compare the efficiency of robot and manual work as well as the costs and savings generated by automation. The profitability analysis of implementing mobile robots is based on several parameters that allow for the comparison of manual operators with autonomous vehicles. The most important of these are working time, labor costs, and work quality. We will present each of them in detail, showing exactly what should be subjected to analysis.
Calculating the actual working time of a robot and an employee
First, we consider the specifics of the company's work that plans to implement robots. Firstly, we take into account how many days a year and how many shifts the plant operates. It is also necessary to consider periodic increases in demand for our services. For example, for a logistics company, it may turn out that the company works from Monday to Friday on two shifts, but from November to December, the work cycle changes to three shifts including weekends. These data should be adjusted for mandatory breaks (meals), periods of morning start-up (briefings), or breaks related to the transfer of duties between individual shifts. It is also possible that there are other technological breaks in the plant when transport processes are not carried out. Finally, we should determine the "net" working time. In practice, where the plant is not fully automated, the robot's work will be closely related to the work of people, which does not mean that it will not be more efficient and predictable.
However, that's not all because determining how long mobile robots will actually perform tasks can only be achieved through pre-implementation analysis. Why is it necessary to obtain data enabling investment returns estimation? We then check how much the work of robots must coincide with the working time of people handling production and storage. Additional parameters are also examined, such as the time of robot implementation into work, real breaks related to battery charging/exchange, and whether vehicles are slowed down or stopped during task execution to avoid collisions with obstacles or pedestrians on the route. If the robot is stopped for only 3 minutes per hour of work, it means two weeks of unjustified break in work per year (in a two-shift system, Mon-Fri). Similarly, manual worker absences must be taken into account, regardless of the reason.
The data to be collected in this area are:
- Number of working days in a calendar year
- Number of working days on weekends
- Number of shifts on working days
- Number of shifts on weekends
- Working time of one shift
- Downtime during shift changes
- Time for breaks related to meals
- Downtime for maintenance work
- Downtime caused by breakdowns
- Unforced robot downtime caused by external factors
- Average time of manual worker absence due to sick leave
- Average time of manual worker absence due to vacation.
Efficiency and quality of work performed by a robot compared to a manual worker
Comparing efficiency requires standardized processes and clearly defined tasks. This could be the transportation of goods from one location to another, such as from production to a warehouse. Measuring the performance of a mobile robot and a manual operator may involve indicators such as:
- Cycle time (e.g. travel from A to B and back)
- Number of pallets transported per hour
- Number of errors made
With this and previously collected data on actual work time, it is possible to calculate the overall efficiency (OEE). This will be the product of three factors:
OEE = availability x performance x quality
Availability is the percentage of time that equipment or an operator is available to perform a task. Performance is the percentage of the maximum theoretical throughput of the process, which is handled by a classical or autonomous cart. Quality is the percentage of journeys completed without losses.
Let's take an example:
OEE manual operator = 0.81 x 0.85 x 0.89 = 61% OEE mobile robot = 0.95 x 0.55 x 0.99 = 52%
Although these are only sample data, it is very possible that you will get similar results. The robot's lower efficiency is evident, which confirms our observations. In internal transport, it cannot be assumed that one operator will be replaced by one robot. Each of them has a different work specificity, and the advantage of a robot is the stability of its work. This means that if we achieve very good results in the optimization cycle, they will be delivered every time, and the experience gained can be transferred without losses to other robots in the fleet of a given warehouse. What level of OEE should we aim for? An OEE result above 70% should be considered very good, and 85% or more is a result at the level of global optimization leaders. On the other hand, any result below 40% should be considered poor and requiring immediate intervention.
Cost of AGV/AMR robot labor vs manual labor cost
Calculating the costs of maintaining a team of operators is not limited to simply adding up their salaries, although this will be a significant value. Especially since in 2023 alone, two increases in the minimum wage are already guaranteed by regulations, which will also result in greater wage pressure on higher positions. It is also important to note that there may be costs associated with people indirectly related to transportation. If robots are to replace a process handled by 6 operators, and tasks will be issued to them by the ERP/WMS system, the position of the supervisor will also become redundant. In addition to the costs directly related to remuneration, there will be costs for OHS training and staff rotation management. Additional savings come from reducing the costs of maintaining a workstation (helmets, overalls, footwear, heating, lighting).
The cost of robot labor includes the purchase price of the device, any software licenses (if they are a separate item in the manufacturer's price list), and maintenance costs. We ignore the costs of electricity required to power the vehicle, assuming that the differences in energy consumption between manual labor and robots are too similar to make a significant difference. However, it is worth noting that with a larger number of AGV/AMR vehicles and continuously rising electricity prices, such a comparison will also be made, as autonomous vehicles more effectively manage movement in the warehouse, performing only the necessary movements to complete a task. The comparison of insurance costs for manual and autonomous vehicles remains to be considered, but since policies vary widely in scope and each company has a different approach to securing its property, we consider this issue to be highly individual.
Data to be collected in this area includes:
Manual workers
- Number of employees x salary level
- OHS training costs
- Onboarding costs
- Cost of staff rotation management (recruiting new employees)
Mobile robots
- Cost of pre-implementation analysis
- Cost of robot purchase and any software licenses (many companies allow the entire or part of the cost of pre-implementation analysis to be deducted from it)
- Additional integrations, e.g. with ERP/WMS systems
- Costs of lower process throughput during the implementation period
During the implementation of mobile robots, there may be other costs that are highly individual, so we have not included them in the main list. These include, for example, the cost of preparing secure wireless communication. Mobile robots use Wi-Fi or 5G networks. According to our observations, connecting them to a company's network can take anywhere from less than an hour to several days. Another thing is that space is extremely valuable in warehouses. While mobile robots are often parked by employees in a tight formation, robots require wider parking spaces to be set aside for them. This is because anti-collision systems prevent them from parking in a tight formation. It is also not uncommon for mobile robots to look almost new after 5 years of use, while those used by manual laborers appear to be at the end of their lifespan. This falls under the category of cost as there are unexpected savings, such as the introduction of robots in the FMCG or pharmaceutical industries, which significantly reduces the cost of managing the risk associated with product contamination.
AGV/AMR in the longer term
The collected data, which shows the current state of affairs, can be used to make forecasts for the coming years. In the case of robots, it can be assumed that after the first year of use, parameters such as speed and overall efficiency may improve. This will be due to people becoming more accustomed to working with robots and optimizations made based on data collected in subsequent months. As for costs, the risk of significant expenses only arises after the warranty period has ended (usually 2-5 years), when additional inspections may require paid parts replacements. After 5 years, mobile robots should have long been fully amortized.
It is also worth considering alternative costs if we decide not to implement robots, such as preparing a simulation of labor costs. Until now, it has mainly taken into account the increasing efficiency that comes with increasing employee experience. However, currently, increasing wage pressure is more often due to inflation and cyclical increases in the minimum wage.
Finally, a useful exercise is to check if the cost of implementing mobile robots could bring us some alternative profit. Would hiring additional workers be more effective or would putting the money earmarked for AGV/AMR into a savings account be more profitable? According to our knowledge, it would not, but everyone must answer this question independently. It is best to do this while analyzing the results of the situation analysis in your own company, which we warmly encourage!
Explanation of TCO and OEE
TCO (Total Cost of Ownership) refers to the total cost of owning a particular piece of equipment for its entire useful life. This indicator takes into account the cost of purchasing the equipment, its maintenance costs, repairs, energy consumption, and depreciation over time.
OEE (Overall Equipment Effectiveness) refers to the efficiency and effectiveness of a machine or even an entire production line. The indicator measures how well a given robot performs its task in relation to its theoretical capacity. OEE mainly takes into account equipment availability time, equipment utilization, and work quality.
In summary, TCO answers the question of ownership and maintenance costs, while OEE indicates its efficiency. Combining both indicators in analysis is very useful because it addresses the initial dilemma of this article: whether despite robots being more expensive than employing a human, their efficiency makes them a profitable investment.
