# Getting Started with Robotics: The First Three Steps for Manufacturing SMEs

Author: Marin Niehues, Niehues Consulting · Published: 2024-08-29 · AI Adoption & Operations
Source: https://www.niehuesconsulting.de/en/blog/getting-started-with-robotics-three-steps/

Getting started with robotics works in three steps: you choose a repetitive, clearly bounded task, calculate the economics on the total investment and launch a pilot cell with named owners. The technology is the smaller hurdle. Whether a robot takes root in daily production is decided by the choice of task, by the calculation and by the question of who runs the cell from the first day.

This article is written for managing directors and plant managers of manufacturing SMEs who want to introduce their first robot. It describes the three steps with a criteria table, a worked calculation and a checklist. System integrators and manufacturers handle the integration of the technology. Which task is automated, how the investment pays back and how the introduction is organised remain decisions for company leadership.

**The first robot should take over a task that repeats every day, has a measurable result and is clearly bounded in the workflow. Anyone who starts with the hardest task in the plant adds months before the first effect.**

## Why is now a good time to start with robotics?

The time is good because skilled workers are scarce and robots for simple handling tasks are already widely used in Germany. According to the International Federation of Robotics (IFR), German industry had a robot density of 415 industrial robots per 10,000 employees in manufacturing in 2022. That is third place worldwide, behind South Korea and Singapore. The comparison shows that competitors at home and in neighbouring markets have gained experience with robots, even though the density is shaped largely by the automotive industry and large plants.

Two developments help mid-sized firms. Collaborative robots, or cobots, work next to people without fixed safety fences where the risk assessment of the application allows it, which reduces floor space and entry cost. Programming tools have also become simpler: many manufacturers offer interfaces with which trained staff set up motion sequences after a short induction, without mastering robot programming languages. Both lower the hurdle for plants with small batch sizes and few sites. An overview of where robots pay off in mid-sized firms is in the article on [robotics in the Mittelstand](/en/blog/robotics-mid-sized-companies-economic-case/).

## Step 1: How do you choose the first task for a robot?

You choose the first task by the five criteria in the table below. Suitable tasks demand the same movement dozens to thousands of times per shift. Palletising finished cartons, loading a machine tool and placing parts into a press are typical. These tasks have fixed parts, fixed positions and a verifiable result.

Tasks with strongly varying parts or subjective quality judgement are unsuitable for a start. They require sensors and machine vision that a first pilot should not solve at the same time. An unstable process also does not belong in the first cell: a robot that grips a part wrongly because the upstream process fluctuates mainly exposes weaknesses of that upstream process.

**Criteria table for the first robot task**
Criterion | Good for a start | Unsuitable for a start | Check question
Repetition | Same movement every shift | Every execution differs | How often does the task run per shift?
Part variety | One to five parts | Dozens of changing variants | How many variants occur in one week?
Presentation | Fixed position, defined rack | Loose bulk, changing orientation | How does the part lie at the moment of gripping?
Process maturity | Stable dimensions, known scrap rate | Fluctuating upstream process | What is the scrap rate today?
Effect | Relieves a bottleneck or a hard-to-staff position | Affects a station without bottleneck | Which metric improves, and by how much?

The selection should not be made by engineering or production management alone. Production and maintenance rate the candidates together with occupational safety, and controlling checks the figures. This regularly produces a list of three to five tasks, from which the one with the best ratio of effect to simplicity starts first. The same logic of rating by value, feasibility and risk applies to any automation.

## Step 2: How do you calculate the economics of a robot?

You calculate the economics of a robot from the total investment of the cell and the annual net relief, which together give the payback period. In practice the robot arm accounts for only part of the total investment. Grippers and safety technology come on top, as do feeding, integration into the line and commissioning with training. Anyone who compares only the robot price with the savings makes the cell look cheaper than it is.

The following example is illustrative and shows the logic. A metal parts manufacturer automates palletising at the end of a line that runs in two shifts.

**Worked example for a palletising cell**
Item | Assumption | Amount
Total investment of the cell | Robot, gripper, safety technology, feeding, integration, commissioning | EUR 120,000
Relieved working time | 1.5 full-time positions moving to value-adding work | 1.5 positions
Cost per full-time position per year | Company assumption | EUR 55,000
Annual relief | 1.5 times EUR 55,000 | EUR 82,500
Annual running costs of the cell | Maintenance, energy, spare parts, software | EUR 8,000
Annual net relief | EUR 82,500 minus EUR 8,000 | EUR 74,500
Payback period | EUR 120,000 divided by EUR 74,500 | about 1.6 years

The calculation holds only if the relieved working time is actually used elsewhere or a vacancy no longer needs filling. If production counts the relief as savings while nothing changes in the shift plan, the amount stays on paper and never reaches the result. The question of where the freed capacity goes therefore belongs before the investment decision. A detailed calculation with floor space and payback is in the article on [planning robot cells](/en/blog/planning-robot-cells-investment-payback-space/).

Four assumptions drive the result most: the total investment, the share of working time that falls away, the utilisation of the cell and the running costs. Check each with the people involved and calculate a cautious case with 30 per cent less relief. If the cell pays back within a period management accepts even then, the investment is robust.

## Step 3: How do you launch a pilot cell?

You launch a pilot cell by settling, before procurement, who operates the cell, how it is maintained and by which metric success is measured after 90 days. A robotics project can reach commissioning and then fail for lack of ownership. If the system stands still and the integrator is only reachable during contract hours, production returns to manual work and the robot becomes an exhibit.

Operations need named roles: an owner for the cell in production, at least two trained operators per shift and a contact in maintenance. Rules for faults belong to this: who decides on which error, and how long does production continue without the robot? The people on the line are involved from the start, because they know the task best and use the cell every day. How teams accept robots in daily work is described in the article on [human-robot collaboration](/en/blog/human-robot-collaboration/).

For applications in which robots and people share a workspace, the safety standards ISO 10218 for industrial robots and the technical specification ISO/TS 15066 for collaborative robots apply. The risk assessment of the whole application is carried out by the integrator together with a safety specialist. This article provides an overview and does not constitute legal advice.

**Checklist before the purchase decision**
- Is the task rated against the five criteria, and is its effect on a metric quantified?
- Is the total investment calculated with all items, including feeding and integration?
- Is it settled where the freed working time goes?
- Does the cell have a named owner and two trained operators per shift?
- Is it settled how production continues during a stoppage?
- Is a success metric fixed for day 90?

## How long does it take from decision to a running cell?

From decision to a running cell, a simple handling task usually takes several months, depending on delivery times and integration effort. The integrator states reliable times in the quotation. From a delivery perspective, what matters most is what happens before the order, because it determines the time after it: the task is defined, the cycle times are documented, the owners are named and the space on the shop floor is released.

A frequent cause of delay is the interface to the rest of production. The robot needs signals from upstream machines and coordination with logistics for feeding and removal. These dependencies should be made visible early in planning. Anyone who discovers them only at commissioning loses weeks.

## Getting started with robotics with Niehues Consulting

[Niehues Consulting](/en/) is a consultancy for AI management and organisation design from a delivery perspective, based in Düsseldorf. Founder Marin Niehues builds organisations that deliver and brings technology into operations, in projects ranging from VC-funded scale-ups to DAX 40 corporations and in more than 20 countries on six continents. The integration of the robot cell lies with system integrators and manufacturers. Niehues Consulting starts before the technology: the business case, the selection and sequencing of applications and the organisation that runs a cell from the first day.

The approach follows the Niehues Delivery Model with its three phases. Three of its checking questions matter for getting started with robotics: Which business outcome should change measurably? Is the business case confirmed with real values? Who is responsible for operations from day one? The entry point is a free initial consultation of 30 minutes, which you can book via the [contact form](/en/#kontakt). The scope and terms of an engagement are discussed in the free initial consultation.

**Conclusion**
Getting started with robotics in a mid-sized firm needs a simple task, a calculation on the total investment and a named owner from the first day. The technology is available, and the bottleneck lies in selection and organisation. A pilot that settles both provides a robust basis for the next cells.

**Sources**
- International Federation of Robotics: Global Robotics Race, Korea, Singapore and Germany in the Lead, press release on robot density, January 2024
- ISO 10218-1 and -2: Robots for industrial environments, safety requirements, 2011
- ISO/TS 15066: Robots and robotic devices, collaborative robots, 2016

## Frequently asked questions

### Where do you start with robotics in a mid-sized manufacturer?

You start with a repetitive, clearly bounded task such as palletising, machine loading or packing. The first step is to rate several candidates by repetition, part variety, presentation, process maturity and effect. The technology is chosen only afterwards.

### How much does a first industrial robot cost?

The price of the robot alone says little about the investment, because grippers, safety technology, feeding, integration and commissioning come on top. Reliable figures come from a quotation for the specific task. Calculate the cell with all items before you judge the payback.

### How long does an industrial robot take to pay back?

The payback period is the total investment divided by the annual net relief. It is short with high utilisation across several shifts and with hard-to-fill positions. A cautious case with less relief shows how robust the result is.

### Are cobots better than traditional industrial robots for a first project?

Cobots suit tasks with small batches, little space and occasional work alongside people. Traditional industrial robots have the advantage at high speeds and heavy loads. The task decides which type fits.

### Do we need programmers for the first robot?

For simple handling tasks, trained staff with a manufacturer induction are often enough, because many systems offer graphical interfaces. More complex applications with sensors or links to order data need integration expertise. The integrator clarifies this in the quotation.

### Who should own the first robot in the company?

Production owns it. A named cell owner steers operation, faults and metrics. Maintenance and occupational safety are involved.

### What belongs in a pilot cell?

A pilot cell has one task, one metric, one owner and a 90-day evaluation period. It runs in the real workflow with real parts. After 90 days, management uses the metric to decide on the next cell.

### How do you involve the workforce in a robotics rollout?

Involve the people on the affected line before procurement and clarify which tasks will change. Training and a clear answer on where the freed working time goes build acceptance. Employees who operate the cell find faults earlier than any monitoring.