The challenges of offering Robots-as-a-Service

Around 2018, we began hearing about RaaS (Robots-as-a-Service) as a new business model for robot vendors to extend their services to customers who either could not afford or could not justify investing in full automation. Inspired by the success of SaaS (Software-as-a-Service) in the enterprise software market, the primary benefit was supposed to be that by purchasing a service contract rather than owning equipment outright, automation would become an operational expense (OpEx) rather than capital expenditure (CapEx), thus avoiding large up-front costs and enabling a business to scale their automation in sync with changing business needs. The service contract would include deployment, integration, upgrade and maintenance of equipment as well as associated services such as cloud-hosted software. As a customer, on top of the fixed subscription costs of a support contract, one would essentially pay for hours of usage on equipment; miles run by an industrial mobile robot, for instance.

Sounds good from the point of view of prospective customers. So what happened to RaaS? It’s only 2022, and we don’t hear so much about it anymore.

For most robot vendors, especially startups, building a sustainable business around RaaS has probably proved to be quite a challenge. Robotics is fundamentally a hardware business — ultimately robots interact with the physical world. Their capabilities are defined by hardware more than software — for instance, the usefulness of an industrial mobile robot is primarily defined by the size of the payload it can carry, the speed at which it can carry it, and by such things as safety systems, endurance on single charge of the battery, etc. Pure software applications are not encumbered by such real-world limitations. SaaS may not be a useful model to base a robotics business on, after all. Here’s why:

1. There is no cost to a SaaS vendor in acquiring a new customer for their software application. In most cases, prospective customers sign up for and download a copy of the product from the web; providing a copy costs nothing. For a robot vendor, the cost of acquiring a new customer is high. The upfront CapEx savings for the customer becomes an upfront cost for the vendor. The vendor has to first build the hardware, which costs time and money, and then has to ship it to the customer, which costs time and money. There are additional costs that a typical software house needn’t bother with, such as warehousing to stock components and manufacturing space to build that said hardware at whatever their scale of operation is.
2. A buggy software is still usable software. You can use it until it crashes on you the next time, or you can work around its limitations. A broken robot is usually just broken and unusable. Therefore, customers expect higher standards of reliability from machines as opposed to an app, especially for industrial/commercial equipment where down-time can be expensive. Engineering reliable hardware takes time to build and validate, and takes highly skilled engineers — of the mechanical and electronics kind, and not just software. Hardware is hard.
3. Getting support for a buggy software application is relatively straight-forward. As a customer, one typically raises a bug report or an incident support request on the web, pays for priority support if that’s not included in the contract already, and if all goes well, they have a new version to download or an over-the-air update ready in due course. A broken hardware has to be either shipped to the vendor or the vendor’s service engineers have to visit the customer to fix the problem on-site. All at vendor’s expense, which can be particularly high if the promised response times are short, such as next business day.
4. Assuming that the robots are functional and reliable, the vendor still does not make money unless the customers use them. RaaS has been promoted as a business model specifically to those customers who don’t need robots all the time. For instance, prospective customers such as 3rd party logistics (3PL) companies experience a natural periodicity to their workload at different scales. Mornings are usually busier than afternoons (as buyers place orders for next day delivery), Mondays are busier than Fridays, beginning of the month is busier than the end of the month (when people may have run out of disposable income) and Christmas times are busier than any other time of the year. Therefore, such customers may consider paying for automation only to ease their peak workloads, whilst getting by with a permanent staff on fixed wages and predictable costs the rest of the time. This means those machines may be sitting idle in the afternoons towards the end of the month in their warehouses. That’s expensive equipment not earning an income for the vendor.

Adding to all the complexity, RaaS vendors may have to serve the role of equipment manufacturer as well as systems integrator. Traditionally, automation systems are not sold directly to large industrial end-users but to Systems Integrators who are hired to offer turn-key solutions by bringing together equipment and services from different vendors to satisfy the unique requirements of an end-user. Systems Integrators prefer to buy equipment outright, commission the system and move on to the next one. RaaS is not an attractive proposition for them.

All these factors mean that there are significant challenges to providing RaaS services at a price that makes sense to both equipment vendors and their end-users.

There is a place for RaaS in specialist services and niche applications such as inspection and maintenance of high value fixed infrastructure (oil and gas installations, for instance), underwater robots for inspecting hulls of ships, flying robots for maintenance of wind turbines, and so on. Such businesses are typically led by enterprising engineers who have spent a lifetime acquiring deep expertise solving problems in these domains, thus reaching a point where they are able to create a viable business out of the solutions they have nurtured over time. These businesses tend to be small (at least until they are acquired by larger corporations), and sustainable in the sense that they are able to pay all their bills and wages and survive for a long time. That’s not a bad thing. A business can’t grow forever, but it can remain sustainable forever.