The heart of the 3D cameras of the O3D series is a new image chip based on the PMD technology with which the devices can recognise objects in their spatial dimensions. However, ke NEXT seemed to find the new operating concept a bit more interesting…
You have also integrated a new user guidance concept and new operating software into the O3D camera navigation. What has motivated you?
Mike Gonschior: At ifm, several subsidiaries define their own products. In our area, we have the 3D cameras with the PMD sensors, we have the vision sensors for code recognition, contour verification and sensors for non-contour objects. There are higher-level circles where we discuss things, but you can easily tell which product comes from which subsidiary. We also have four different types of operating software that have drifted more and more apart. Even with regard to hardware operation you can see that different operating elements and different setting options are used for products.
This is why we have decided to develop an entire department dedicated to user experience and to hand the user guidance development over to this department. Our software, the Vision Assistant, is now the result, and the O3D cameras are the first product series designed according to this model. It is a platform where several products will be operated in the future.
Was it necessary to form an entire department for this?
Mike Gonschior: Unifying the operation of the products was not the only objective. Apart from that, we have defined it as an important goal for ifm to develop a significant competitive advantage via usability. This is indeed a main topic for us. This is why we have a new department that pushes this. This department’s aim is to ensure in the long term that handling our products, even if they get more complex, stays easy – and this across all product groups. The user does not need to constantly learn handling new operating concepts, which we consider to be a great competitive advantage.
Why is easy handling so important?
Mike Gonschior: Why was Apple’s smartphone such a success? The reason for this was indeed the user guidance. A smartphone is indeed a miracle weapon. One person uses it to call people, the other person uses it to read books, the next person uses it to write e-mails or as a camera, another person uses it to scan business cards. However, basically, no one is capable of handling such versatility. This can only work if this versatility is structured in apps, i.e. in clear application-specific reduced handling. We envisage the same principle for our products. Human experience is not only based on our working environment, but you also get an extreme amount of impressions from consumer products. This is why everyone’s expectation is also raised when it comes to industrial products.
Karolina Beck: Exactly: A development process focusing on the user requires exact knowledge about how the users actually will handle products in the end. What is the context in which the user works? What are the tasks? How does the user handle occurring problems? And what is done to solve them? In order to know what easy handling is, it is also important to know for whom which handling is easy – and why. Who is the target group, who is a persona? In our case, we have defined three personas. One of them is the specialist, the image processor and PLC programmer. The second persona is the maintenance technician, a generalist who wants to solve his application quickly and easily. And the third persona is our own sales engineer. While an image processor needs a complete overview of all functions for set-up operation, a maintenance technician would rather like to reach the goal in five steps.
How did you actually find out what the user really wants? The industrial sector is relatively heterogeneous. There are many different approaches.
Karolina Beck: The user experience process is subdivided into three stages. There is the analysis stage, the conception and the implementation. During the analysis stage, methods from ethnography and psychology are applied. Often these are interviews. We have conversations with the users. We do not ask superficially what they want, but what they are doing with and what their goal is. Another method is simple observation: How does this person use the system? Sometimes it is rewarding to spend an entire working day with the user in order to get a feel for what is important and what is less important. You watch the daily routine and how problems are solved.
Then, the mentioned personas are created.These are models of typical users that make it possible to develop products with a greater feel for the user. After that, we create scenarios: In which context is the product used? What are the user’s task, goal and requirements? You simply cannot compare using the product when you have time and can make all settings without any stress to an emergency situation in which the machine operator is looking over you shoulder, waiting to resume operation of the machine. This is a completely different context which requires a completely different handling. Afterwards, interaction concepts are developed and tested in the scenarios. At the end, prototypes are developed and usability tests are carried out based on the defined personas.
Does the customer know what he wants?
Mike Gonschior: You really have to concentrate on the customer and read between the lines. Nobody said they wanted a larger screen when the Nokia phone was released. Everyone thought they needed a smaller phone. Nevertheless, the iPhone became a great success. The truth is: People always find a work-around. If there is a mouse placed right beside the modern touch screen, the user may not even identify this as a problem. But if you see this mouse, you should ask what the user needs it for. And if it turns out that the software can be improved to make the mouse redundant, you should do that.
Why are users dissatisfied with software?
Karolina Beck: For example because they do not see the scroll bar because the contrast is not high enough. At a certain age, the sensitivity for contrasts begins to decrease and you might not be able to recognise that there is a scroll bar. This is particularly unfavourable if the important information is situated in the lower part.
Mike Gonschior: Another example is our wiring wizard. We realised that wiring is a challenge for some customers. For standard 5-pole cables, the ground or zero is usually blue and the voltage supply is brown. Nevertheless, there are connectors which deviate from this standard. Therefore, we decided to offer a wizard that helps to avoid faulty wiring. This solution is quite simple, but a good example for user-centred design.
What has surprised you in the context of your research?
Karolina Beck: You always have to expect the unexpected. Everybody has their own mental model of things. I had to learn that the user, particularly the end user, the operator of the machine, has a different way of thinking than the developer.
Industrial products are extremely feature-based. That means the customers usually express their requirements in the form of technical specifications. They do not provide specifications regarding usability. Besides, communication usually takes place among technicians from both sides. No one asks: How do the users feel when they use the component or operate the machine?
The feedback we receive during testing is also very heterogeneous. For some users, for example, it was clearly a problem when the text shown in a wizard was not ideal. That means they really read the texts. Others tended to click through the wizard without giving much attention to the texts. That surprised me as well.
What is the greatest challenge when implementing such a process?
A wizard is also used for the teaching process – in the example completeness monitoring of crates. Among others, the user is asked whether crates in an inclined position are acceptable. Image: ifm
Mike Gonschior: I think the most difficult aspect is that user guidance will always be a compromise between standardisation, that means similar handling of as many products as possible, and optimisation of individual products. When new products are created, you will always have to make a decision: what degree of standardisation provides maximum benefit to the users of different user groups.
How have you integrated all of these criteria into the operating software?
Karolina Beck: We started with the set-up of the software. Specialists are provided with all setting options and can test all parameters of the O3D at their leisure. Generalists need a fast, compact solution. They select their application and are guided through the wizard in only six steps. We also adapted the design. Industrial software often looks a bit like Windows 3.1 and cannot be compared to modern touch screen software surfaces of smartphones or tablets.
We made sure that the software can be operated via touch screen, because we found that such tools are more and more used in the field. We also took into account colour blindness and that the software may be presented via a projector. In this case as well, the software has to function reliably and the font has to be readable. We optimised all these technical basics.
How did you know you were on the right track?
Karolina Beck: We repeatedly carried out tests. In a first step, internally, then with customers. The test users received a task they had to solve using a click dummy. We observed whether they could connect the product, whether they found the start button, whether they understood the wizard correctly and whether they were lead to the correct interactions. Then we further improved this interaction based on the user feedback.
In a first step, the new software will be available only for O3D cameras. Did the new usability concept also influence the design of the camera hardware?
Mike Gonschior: Of course, for example regarding the position of the connections. Besides, we made sure that the LEDs are visible from different angles. Even without a display, we are able to provide a clearly visible feedback to the user of what the product is currently doing via different flashing signals.
Why did you omit the display?
Mike Gonschior: The main reason is that the O3Ds are mainly used as sensors. They are often installed in locations where a display can not be seen. Colour LEDs can be seen more easily from a distance. For details, there is still the software.
Let us discuss the camera in general. What is special about the 3D sensor concept of the O3Ds?
Mike Gonschior:
The camera uses PMD technology. I would like to elaborate on this.In 2000, ifm founded a group focusing on industrial image processing, developed several products and bought technical know-how. Among others, ifm took a stake in PMD Tec, a spin-off company of the University of Siegen. Meanwhile, the company has become a wholly-owned subsidiary of ifm. PMD Tec mainly produces photoelectric sensors whose function is based on the time-of-flight principle for the consumer, automotive and industrial market.
In 2005, we launched the first product based on this technology and immediately won a Hermes Award In 2009, the first PMD camera with a resolution of 3000 pixels was launched. The current cameras of the O3D series have 23,000 pixels. Actually, they even have 100,000 pixels, but we merge four pixels each to improve the measuring accuracy. The distance to the camera can be calculated in real time for all 23,000 pixels, so that it is also possible to measure moving objects. When customers realise that even persons within the detection range of the camera can be represented in three dimensions, this inspires creativity. This is why there is such a broad range of applications.
What to expect in the area of 3D sensors?
Mike Gonschior:
Meanwhile, special consumer chips from PMD Tec are used in Asus and Lenovo smartphones. They make it possible, for example, to suppress the background while making video calls. Based on these sensors for the mass market, we will launch a mini 3D camera for the prosumer market. This includes for example cleaning robots or robot assistants for hotels or hospitals. Such devices often still contain consumer sensors such as the Kinect camera – we will place a professional, cost-effective alternative on the market.
ke NEXT background knowledge
High-resolution 3D camera O3D captures objects in real time
With the new O3D sensors, ifm have developed an innovative generation of 3D sensors.The 3D cameras based on these sensors see various scenes and objects in their spatial dimensions at a glance. The O3D image chips are PMD (photonic mixing device) sensors, i.e. photoelectric sensors whose operating principle is based on the time-of-flight principle. In contrast to laser scanners which only have one receiving element the image chip of the new devices features 23,000 pixels. Therefore the 3D cameras have no moving parts, making them wear-free and robust.
Output is done via grey-scale value and distance images, evaluation is easily possible using common libraries. As each individual pixel measures the distance it is, for example, possible to precisely dimension and position suitcases, parcels and pallets. The devices make it possible to determine and compare dimensions, orientation and position. Moreover, the system facilitates automatic loading and unloading of containers and storage room optimisation, as the camera provides height, width and length to calculate strap length and volume. Besides, the completeness of packages and crates can be monitored. Other applications include collision avoidance, navigation of driverless transport systems as well as counting of people and room monitoring. Even trunk measurement in a saw mill poses no problem. There is a wide selection of applications in different industries.
About the author
Wolfgang Kräußlich – ke NEXT Editor-in-Chief
Wolfgang Kräußlich studied Physics and Forestry at LMU Munich. He has been working as a specialised journalist since the early 2000s, focusing on mechanical engineering and automation.
Phone.: +49 8191 125-667
Threedimensional detection and analysis with the O3D sensor by ifm. Do you want to learn more about the O3D?
ifm is a manufacturer of sensors, communication and control systems. After many years of intensive cooperation with our customers we have established ourselves in the market as service-oriented sensor specialists and today we are represented by more than 6,700 employees in over 70 countries worldwide.
Picture credits:
© ifm electronic
© pmdtechnologies
