Artificial Intelligence (AI) and wind turbine blades, part 2 

We, at Bladefence, strongly believe that automation and robots will play a significant role in the future of blade maintenance.

This is the second part of a two-part series, where I will be looking into how AI and automation are impacting wind turbine blade maintenance. In the first part, published in July, I wrote about blade inspections and the asset management aspect of it. In this post, I will be looking at automated blade maintenance systems, the recent developments, and the outlook in this field. 

Over the last several years automated blade repair systems have been introduced at an increasing pace. Multiple players have introduced concepts and systems to tackle this problem. All of them have a common goal; to reduce costs, improve blade reliability and, of course, increase safety by reducing the requirement for humans to directly work with blades at height. We, at Bladefence, share this vision and strongly believe that automation and robots will play a significant role in the future of blade maintenance. 

To understand the current situation and the steps needed for automated systems to fully breakthrough in blade maintenance, we first need to define what automated blade maintenance means. A good comparison is the “SAE J3016 Automation Levels for Motor Vehicles”. In short, this SAE classification, first published in 2014, defines the role of the driver, rather than technological capabilities of any systems installed in cars. Level 0 is no automation at all, and Level 5 is full automation where the vehicle systems are managing all vehicle and driving functions, including 1) direction and speed control, 2) monitoring driving environment and 3) the system has full fallback responsibility. In other words, in Level 5 humans are completely removed from the equation. 

Over the last decade the automotive industry has poured significant R&D resources into implementing self-driving systems into their vehicles. And we are starting to see the results. Back in 2016 Tesla announced that all of their vehicles would be shipped with the hardware necessary for “full self-driving”. Tesla CEO Elon Musk promised that “you will be able to nap in your car while it drives you to work”.  

I have immense respect for the ambitious goal Elon Musk set for Tesla and I am sure they will one day deliver on that promise. However, the reality is that over the seven years since that announcement, Tesla Full Self-Driving (FSD) is still in test mode and is considered to be on Level 2 on the SAE classification, meaning that full human supervision is always required and the driver is assumed to be in full control, rather than system, even when the systems are engaged. Do not get me wrong, even in its current form the Tesla FSD is super-impressive and a testament to the hard work and ingenuity of the Tesla engineers. It is just that the road is long and very difficult

Much like automated blade maintenance. 

To reach the primary goal of cost reduction, the automated blade maintenance systems will be required to accomplish at least two key objectives: 1) reduce the amount of people required for blade maintenance and 2) increase speed and efficiency. In combination these will reduce the unit costs and enable wide-scale adoption. The increased safety and quality will be important too, but not deciding factors. How do we know this? Bladefence has used aerial work platforms for blade access for over a decade and in terms of safety, speed, and quality they are far superior to rope access, but the market clearly values the unit costs more and has leaned towards rope and suspended platform access. We have adapted our offerings to suit this market demand as well.  

This sets the stage and overall requirements for automated blade maintenance systems – they need to be as autonomous and as efficient as possible to reduce the unit costs significantly. If the same or similar amount of personnel is still needed on-site to monitor and control the systems and the speed of repairs is not significantly increased, it is going to be an uphill battle. 

I am not saying it is impossible, far from it, it is just exceedingly difficult, and the problems ahead will likely be similar to what Full Self-Driving systems are currently dealing with. 

To make things even more difficult the wind industry remains a very closed environment where information sharing is non-existent and 3rd party development is that much more complicated. In many cases the only option for an external system developer to make any money is to be bought out by an OEM. There are, however, many successful examples of 3rd party companies entering the blade inspection market with drones, but the difference between inspecting blades and repairing them cannot be understated.  

As a forever optimist, I remain confident that automated systems will help wind energy to become even more cost effective and reliable in the future. 

Ville Karkkolainen

Chief Business Development Officer
August 31, 2023