I believe the ROV industry is entering a new phase, because it has been a nascent and rather small industry. It has, since its beginning, been in a phase of coming up with solutions based on how easy they are to design and develop. In a small market, you are not going to sell many of these items, so it’s not a good idea to invest a ton of money in designing them.
Now, we are entering a phase, not based on ease of design, but based on ease of use because it is a real industry now. This is a cycle that many other industries have been through.
Product design has a really big influence on how easy something is to use or not, whatever the product is. One of the big challenges is that it is incredibly difficult to design things that are easy to use. By corollary, it is really easy to design stuff that is very hard to use.
Ease of use in design
If we look at a familiar item, the iPhone, it is an amazing item when you think about all the utilities it has. It has messaging, you can take photos, you can take videos, it gives turn by turn voice navigation, it has weather, an alarm clock, etc. This is a very clever machine under the covers, but the designers have been able to make it very simple and easy to use on the surface with no need for training.
A less familiar object is the unicycle. This is a personal transportation device that takes a fair amount of training and practice to ride. Add an additional wheel and it becomes easier to use, but still takes a bit of practice.
Another example is an automatic electronic defibrillator. You open the lid, place the electrodes on the patient—and there are pictograms to tell you where to put them on the patient—and then you press the button. That’s all you need to do in order to operate this life saving device.
If that’s what the commercial version is like, what is the professional version like? An excerpt in a manual on how to use a professional defibrillator talks about how to shock the patient, and then has a very interesting caution message in the manual. It says “only use thumbs to depress the shock button. Failure to do so could result in an inadvertent depression of the energy select button, causing the defibrillator to disarm itself”. So, if you don’t use the right digits on your hands, you might get the exact opposite effect out of this machine than the one you had intended. That is bad design. Even worse, the solution to the bad design is only written in the manual. It requires more training than the commercial version that you and I can buy for US$1000.
The automotive industry is another good example of the evolution of focus in design on to ease of use. On an automobile from around 100 years ago there were a number of special things you had to do just to start the vehicle, let alone operate it. Fast forward to 2013, and interfaces on the controls of cars have got to the point that there are buttons that just have your intent on them—“start car”. Ease of use is a big deal.
Designing ease of use into ROVs
An automobile, at a block diagram level, is about equivalent in complexity to a work-class remotely operated vehicle (ROV) system. The difference is that automobiles have been around a lot longer, and the industry has worked on a lot of ways to make them very easy to use.
An interesting benefit of ease of use, and this is particularly relevant to the phase that the ROV industry is now in, is that it is easier to train people on things that are easier to use.
It is also very common for someone to invest millions or tens of millions of dollars in a piece of equipment and not have budgeted into it the fact they want their staff to be competent on it.
At Schilling, we have been making manipulator arms for many years. They may look really easy to use in the hands of a skilled operator, but they are actually really difficult to use. What makes it so difficult is that the manipulator arm has six degrees of freedom (six joints) on it. To grab a T-handle, properly and gracefully, each of those six degrees of freedom have to be in a unique position at the same time. It’s an exceedingly difficult thing to do without training.
One of the things we have been working on, is a way to have a shared control relationship between a machine and the user, recognizing that the most important contribution that the user makes is actually to supply the intent – what they want the machine to do. The machine then takes care of all the other complicated control pieces.
One of the utilities we are getting ready to release does that. It has a camera on the end of the manipulator, which you point at a T-handle. It finds in the view what it thinks is a T-handle and indicates to the operator by wire-framing it, asking “is this the handle you want”. The operator says yes, and then the machine grasps the handle very precisely and gracefully for the user. This is a shared relationship, with the user supplying intent and the machine doing all the complicated work.
When you grasp a T-handle that way, it is also a lot easier on the manipulator and the tool. The forces are much lower and we are predicting equipment will last longer. Another issue related to these tools, and other objects, is that they are not always stationary. What we are doing in that instance is very similar—the manipulator can track an object.
So, in addition to an automatic handle grasping routine, a second routine we are perfecting is where the cameras on the ROV allow the system to determine what is the “normal” coming out of the panel.
Our tests, so far, are yielding really good results. I can sit down with someone with very little or no training and show them how to pick up a hot stab and stick it in a receptacle, and they do it accurately every time.
 |
| Tyler Schilling |
Service and repair
One side of this story is about how you run these machines when you are doing a job. Another is what do you do when you service and repair these machines.
We utilize a philosophy to design things that can be serviced or repaired within 60 minutes. From the beginning of our company, we really thought we were great at making things serviceable and repairable. We would have endless debates what direction the screw should come in from, because if we did this it would be too small for your fingers and so on. We were looking at that kind of atomic detail.
So, we put a requirement on the design of new equipment that you have to be able to service it or repair it in less than 60 minutes. With some existing designs you can’t actually open safely and close the container within 60 minutes, let alone put your hand on the component you are interested in.
This led to a design philosophy where service components have to be on the outer layer of the machine’s architecture—rather than buried inside. An example of that is a power supply on one of our newer designs. This is held on to the system with three fasteners on the device. It attaches to the system by what we would call a kinematic mount. A coupling guides the mating of the device on to the system, so you can’t do things like pinch O-rings, bend electrical pins, and do all the things that show up in a training course not to do. This means you cannot actually do it wrong.
Some of the philosophies that we have developed in thinking about ease of use is trying to be obvious, honest, and simple. That means there have to be buried complexities. This goes back to the iPhone. Due to clever design development, it looks very simple on the outside.
Bastions of bad engineering
On our new designs, if you want to be fired, design something that needs calibration. This is one of the last bastions of bad engineering.
We now, in our products, artificially force three iterations of a design. We design and build one. The engineer thinks this thing is wonderful. We then force him/her to redesign it and build another one, and then we force a third build.
The funny part is, every single time the next generation item gets built, nobody can even bear to look at the original one that they thought was so wonderful. The reason for this is that modification of design is infinitely easier than synthesis. It is expensive in the development process, but it saves so much time and money for our customers in the field.
The premise, again, is that product design profoundly influences how easy a product is to use. Other industries have been through the same thing—aviation, telecommunications and personal transportation—they are just way ahead of us.
By Tyler Schilling, ROV Conference, Aberdeen, September 6, 2013.
Subscribe
