How to build a world scale ready robotic fleet

Jennifer Gill Roberts
8 min readJul 28, 2023


A major evolution in the deep tech industry has been a shift away from investment focused on the strength of a company’s IP and technology team towards the ability of a given company to create a robust platform, business model and unit economics that drive revenue, scale and profitability. Customers are looking for companies to deliver platforms that have high availability and utilization, deliver operational insights, and ensure long term operation.

In this blog I’ll dive into what it takes for a robotics company to get to this level of world-scale ready. It’s a tough balancing act. Sophisticated fleet management models are needed to execute robot supervision, track performance and utilization, while managing a profitable service model that maintains the fleet through retrofits and software upgrades. All of this makes or breaks a company’s ability to scale and establish market leadership.

In June I invited Arno Strotgen to present to Grit’s portfolio CEOs how to grow a successful service business and also to share his experience as an advisor to several robotics startups. Before starting his journey in the startup world, Arno worked for 28 years at ABB Robotics, most recently as Group Senior Vice President, Head of Customer Service & Digital Platform for ABB Robotics and Discrete Automation. Read on for four key takeaways from Arno and some insights from managing Grit’s portfolio.

It’s not even the first inning

First, let’s keep in mind that we are at the beginning of a twenty year AI-powered automation wave. Demand is seemingly infinite. New industrial opportunities, re-shoring manufacturing, high-mix/low volume production, and technology advancements are all driving explosive growth in robotics. Industrial Robotics operation stock grew by 14% CAGR from 2016–2121 with 517,000 new robots deployed in 2021 reaching an installed base of robots of 3.5 million units.¹

Labor shortages and wage increases are driving growth in robotics globally. Wages in China have doubled since 2006 and China is now increasingly outsourcing manufacturing to Vietnam. It’s not surprising that over 50% of all industrial robots have been sold in China. The US will have 4.6M manufacturing jobs to fill between 2018–2028 and is investing heavily in automation.²

“Manufacturing productivity will surge with wider adoption of robots, boosting output per worker by up to 30% over the medium term.” — Arno Strotgen

We still have a long ways to go. The top robot to human density ratio is in Korea with 766 robots per 10,000 employees. The United States is considerably lower.³

Moving from caged to uncaged robots

Initially industrial robots were mainly present in the automotive industry, due to high volumes of highly repetitive simple tasks. Today the automotive and electronics industry sectors are still the biggest users of industrial robots. The highest growth rate, however, is coming from new applications in small & medium size companies. In 2021 collaborative robots (cobots) grew approximately 50% due to the ease of use and the capability to work alongside humans without the safety fences.

Top use cases for cobots include assembly, machine tending, quality inspection, material handling, packaging , palletizing, gluing and welding. Service robotics are also starting to take off in cleaning, maintenance and inspection.

New use cases are also opening up. Cobots now support higher payload and longer reach enabling heavy industrial applications like big yard logistics and construction trades. Transition in manufacturing to high mix/low volume requires increasing flexibility at the same time Generative AI applied to the physical world is enabling faster robotic learning to meet this need.

Large customers in every sector are eager to deploy innovative new solutions to address increasing labor costs. Startups and major players are also benefiting from lower fully assembled costs which support improved unit economics and profitability. Hardware and enabling software prices are expected to drop by 20% over the next decade driving faster ROIs. Notably, sensors have become more powerful and less costly. Increasing acceptance of Robotics-as-a-Service (RaaS) provides flexible business models for startups and customers.

The world is ready but startups have a mountain to climb in delivering reliable high performance low cost robots.

World-scale ready demands reliability as well as performance

Customers are initially focused on performance as defined by their unique KPIs including cycle time (time to complete a specific task), throughput (for example, typically number of operations per unit time), utilization (number of hours the robot is in operative relative to the maximum possible) or efficiency (percentage of time robots performs productive work). Equally important is durability in the field and high availability which may require repairs, retrofits and continuous software improvements.

Take-away #1: Service can and should be a profit center — whether you are selling the robots or delivering it as a service.

Early stage startups often focus entirely on performance only to realize as they start deploying them how important it is to establish a service-based business model.

Arno shared a thoughtful framework for companies to think about the lifecycle of their product/service.

The customer journey drives this service model. Suppliers need to stay close to customers to generate insights for the next investment cycle and to solve quality and performance issues. But it’s important for them to know where they are in the Value-Added Life Cycle Services model which is defined by four phases.

1. Startup phase: focus on getting the customer up and running. Engineers are likely to still be involved in debugging and training.

2. Production phase: focus on living up to the Service Level Agreements (SLAs) including availability, cost control, and performance metrics.

3. Production improvement phase: Focus on driving operational efficiency with product improvement, for example improved productivity, reduced cycle time, increased production input.

4. Lifetime extension phase: This phase starts a few years before the end of the robot’s lifetime. Do you offer replacements? Refurbish? It may be hard to procure and maintain spare parts. Or maybe you just upsell them to the latest robot. Sometimes a refurbished robot costs more than a new one, but it might be the best for the company that already has the same models and doesn’t want to invest in training and spare part stock.

Take-away #2: Invest in tools and infrastructure to manage the customer journey

“Having a good system to track your installed base is something very critical that a lot of people forget about” — Arno Strotgen

The average lifecycle of a robot is 12 years. What is a successful journey for your product? You will need to manage:

  1. Spare parts
  2. Technical support
  3. Field operations and training
  4. Service Level Agreements
  5. Upgrades and replacements
  6. Optimization

Companies will need a scoreboard or tool to support the process. Arno dove into a few of these topics.

Take-away #3: Develop a plan to manage service for the lifetime of the robot

Spare Parts is a science unto itself. It’s all about having the right part at the right place at the right time. How do you estimate how many components you will need during the lifetime of a robot and how do you price them? Which components will fail the most. You need to guarantee your customer that you will have spare parts for the lifetime of the robot. It’s important to do an analysis on perceived value vs. competition to understand how to manage each category.

Parts can be divided into groups:

Key parts: Higher cost multipliers. For example: motors, drive units, end effectors. Might be specific to your robot. Material price to sales price = 8–10X. Company needs to acquire and manages inventory.

General: For example: cables, electronics. Easy for customer to procure.

Commercial: The company procures, but cannot charge as much

Standard: The customer normally procures directly from the market

Company’s need to create a model and do the analysis. Take the BOM and determine what is the value perception of the customer. What is the risk of failure. For example, the risk might be 20%. Knowing your spare parts model, it’s also important to include service for components with the company’s suppliers.

Field Operations

It’s all about availability! When customers calls, how fast is the response time? How do you capture information about this event into a system for later use? How do you measure customer satisfaction?

“Field operations is all about having the people with the right know how to support your customer via phone or on premise when it is needed to assure the best performance of you customer robots. It requires to have optimal processes, the right tools and spare part availability.” — Arno Strotgen

Support tools, such as remote monitoring, are becoming more important to predict failures and solve issues faster.

Take-away #4: Understand and measure your Key Performance Indicators. Memorialize them in Service Level Agreements

Let’s start with the premise that the lifetime of an industrial robot is typically 12 years and Collaborative Robots (cobots) should aspire to the same lifetime. A lot of companies focus on the upfront cost of the robot and forget that during the lifetime spare parts, training and maintenance need to be included.

“It is important to run service as a profit center, and not to see it as a cost center.” — Arno Strotgen

In order to turn service into a profitable business it’s important to invest in a marketing and sales team for service and to have dedicated product management. Also critically important is the sales of Service Level Agreements (SLAs). An important KPI to measure is how many robots you have under SLA compared to your total installed base. SLAs will give you a steady income stream and also increase customer satisfaction.

“In the later phase of the product lifetime, more than 30% of revenues will come from service. And overall service is a big part of company profits” — Arno Strotgen

What about Raas?

Robotics-as-a-Service (RaaS) is a business model that is gaining in popularity due to the many benefits. The ability to gain access to the latest technology without a major upfront investment and the flexibility to scale your business depending on market demand are only two of the important benefits. Additionally the model provides transparency as the customer will know up front the total cost of ownership including all service and maintenance cost.

Although it can be a very attractive business model for customers and providers it has certain challenges that must be addressed. If you are a provider of RaaS, you have to be certain you can maintain your equiprment throughout it’s lifetime and be able to mantain good cost control over the maintenance and upgrades needed. You also need to know that your customer is willing to invest in training of their employees so that they can operate your equipment according to your specifications.

“Don’t tie the business model (RaaS) to factors you don’t control at the customer site.” — Arno Strotgen

For customers, RaaS makes sense if you require the flexibility and scalability and if the contracted process is not your core business and the provider can offer an overall improved cost of ownership.

Time to scale!


¹ source: World Robotics Industrial Robots 2022 Report

² source: Deloitte Robots on the move: Professional service robots set for double-digit growth)

³: IFR Robotics Report in 2023 What the Latest World Robotics Report Tells Us)