How are technical documentation, eLearning, and assessments connected in your field of work, and what advantages does personalization offer in this context?

Martina Schmidt: We are predominantly active in the technical information sector, but not exclusively. That is, we develop information as technical documentation or eLearning for service personnel so they can operate technical devices safely or maintain them in a safe operating condition. Safety thus has the highest priority.
In this connection, technical documentation is the mother of technical information, which is developed by means of risk analysis. It is important to keep in mind that eLearning to train service technicians’ competence (competence being the sum of knowledge and skills, as well as attitude and awareness) is developed from suitable didactic support materials - and only from the documentation. We already consider this in the documentation’s granulation and structure and have invented a method, the iLEARN, in which we again deploy the learning-objects approach.
Learning objects consist of text and media, as well as an appropriate integrated assessment strategy and mandatory demonstration of prior knowledge. We use a mixture of the DITA Learning Object, the S1000D Learning Data Module, the CISCO Reusable Learning Object, and ideas and media extensions we have developed. This guarantees that eLearning and technical documentation are consistent and that their content doesn’t drift apart. It also makes the design of individualized learning content and non-linear learning paths possible.

But in regard to security-related topics, how can I ensure that competencies have actually been acquired? Knowledge is easy to check with tests; skills or attitude, though, are a different story.

Dr. Harald Stadlbauer: We develop the learning objectives from a combination of learning OCRs (objectives and key results) with the digital variant of Bloom's Taxonomy at the verb level. Using it, the media and interaction strategies necessary to train the competencies can be planned at the same time. The assessment strategies are produced in the process automatically, and with our method, the xAPI statements are derived simultaneously. Thus, via standard cmi5 interfaces, a legacy LMS can be filled and used. It sounds clear and simple, doesn't it?
 

You work in predominantly technical industries - from mechanical engineering to railroad technology to medical technology. What additional benefits can be generated through the deployment of virtual and augmented reality here?

Barbara Kalous: In these technical fields, it's always a matter of absolutely guaranteed competencies in order to be able to handle and operate a product, a machine, or a piece of equipment safely and without danger, and in the training process, a distinction has always been made between knowledge transfer and skills transfer.

Whereas knowledge transfer was gradually provided digitally - as eLearning, training videos, or a combination of the two - skills transfer was always done in classroom courses in training centers and using hardware simulators that had to be created at great expense.
During the coronavirus pandemic, though, this was no longer feasible, and the time had come to try out other technologies such as virtual or augmented reality. This gave rise to
flagship projects such as KONE's VR training, whose amazing effects have led to it now being implemented everywhere.

However, in areas where security was critical, the problem of certification remained. How can I integrate the performance data from VR and AR training environments or simulators into my HR software or my LMS?
This was the moment where xAPI had to be combined with cmi5. Every LMS can implement cmi5. Through xAPI, individual skills can be captured, and with cmi5, they can be aggregated in the LMS.

Which of your "best practices" can be characterized as role models?

Martina Schmidt: Oh, this raises the question of what it takes to be a role model. Isn't that word a bit too grandiose? At NINEFEB, we have the vision that learning comes from the workplace. For example, on the topic of performance support, technical documentation, e.g. diagnosis, meets eLearning for simple training contents and instructions. Technicians will remember anything they learn during the troubleshooting process for the rest of their lives!
Our vision encompasses three specific points:

1. Give technicians multimedia instructions on how to solve a problem that is currently being faced.
2. Give technicians the opportunity to select and consume knowledge content as they want, as long as they have the foundation (non-linear learning paths).
3. Give technicians the opportunity to choose the medium they can deal with best (VR, AR, animations, simulators, images, or text), so they are able to retain whatever they have learned through the convergence of the various media.

Through the appropriate granularity of content - the appropriate selection of media with the aid of Bloom's digital taxonomy - the linking of all these issues to xAPI taxonomies in the iLEARN framework is easier than it might appear at first glance, and the links to competency-management systems can be closed in the process.

A topic of special significance for us is the assessment of training’s impact on the general goals it is supposed to fulfill. In our industrial-engineering world, a business goal such as "service must be faster", "a zero-defects policy", etc., is usually what sets a training activity in motion.
At this point, it is important to plan the training in light of the objective and, when it’s finished, to be able to ascertain whether the training measures have actually achieved the business objective.

Dr. Harald Stadlbauer: As a starting point, we have taken Google's Objectives and Key Results (OKRs) target system and designed the learning OKRs that link the business objective with the learning objectives, with the learning objectives being formulated on the basis of Bloom's digital taxonomy; however, additional measurable Key Results are also defined. These Key Results are mathematically aggregated upwards to the Key Results of the business objectives, providing an immediate monitoring process and the connection to the actual goal. In our learning OKRs, however, the link to xAPI statements is also evinced, which are then automatically converted so that the details can also be gathered, e.g. by measurement, in relation to the respective OKR goal.
One of our models is the Total Learning Architecture of the Advanced Distributed Learning Group (ADL). We see that with our approaches, we are able to solve problems that are currently open in the industry and can do this without very expensive investments involving the maintenance and use of legacy systems.
We have been evolving steadily. Once upon a time, DB Schenker had a type of logistics ERP system programmed that was to be rolled out internationally. We were commissioned to develop context-sensitive help, technical documentation, and eLearning for two different roles, as well as knowledge management for FAQs. Here the focus was more on efficiency; non-linear paths were definitely not an issue. What was interesting, however, was how quickly - using a method of ours called Semantic Content Single Sourcing - we were able to generate the context-specific content from a comprehensive content pool. The video training was limited to screen content.
Through intensive cooperation with an elevator manufacturer and a machinery producer, we have steadily been involved in the realm of non-linear learning paths for about three years. As is common in our industries, there are four to five competence levels that build on each other. These constitute an ideal model, but often have little direct connection with the service technicians’ actual practical experience.
Nowadays, identical courses for new products or new initiatives are offered for all competency profiles in which a veteran sits next to a rookie. What a waste of resources!

Scenarios in which individuals have no relevant practical experience and have to supplement their knowledge also present opportunities for using non-linear paths. Suppose I start out on a job that always involves operating the same type of machine. At some point, I know this machine inside out but have forgotten everything else (see Ebbinghaus’s “Forgetting Curve”). In this case, it would make no sense to go back to training for that machine, even if it is at a higher level. All this can be captured and managed to a large degree by xAPI.
 

In your opinion, how important is the use of personalized learning in technical companies?

Barbara Kalous: Let's go inside the companies. The companies and their after-sales areas, as we know them, either have external service partners or internal service structures. Both after-sales structures, however, represent the company and its product quality.
Time devoted to learning, though, is always associated with a loss of sales. This is why the openness of local managers of the service units, who are evaluated based on sales, is rather modest towards long courses, whether face-to-face or eLearning, and so employees are not trained. Personalized learning can be much more easily consumed in small doses or bits, but also viewed directly in front of the machine.
We are supporters of microcredentials or badge systems here, so I can collect badges and then achieve the next competency status upon completion of all courses.
Clearly, personalized learning has a direct impact on the learners as well. Knowledge gaps can be closed easily. And if they are no longer presented with something they already know, the learning employees have a perceptible added value. Generation Z in particular is not prepared to waste time on useless things and wants to devote its loyalty to meaningful activities.

Only personalized learning turns the company into a learning organization.