Here is a letter I sent to the VP in charge of Pro/Engineer development.

Mr. Michael M. Campbell

Vice President of Product Management, MCAD Solutions

PTC Corporate Headquarters

140 Kendrick Street

Needham, MA 02494 USA

November 22, 2007

Re:  Reply to your letter from January 5, 2007

Dear Mr. Campbell,

Thank you for taking the time to reply to my letter personally.  I disagree with you on your claim that ProE does not cause or aggravate mouse-related RSI (which I will henceforth also refer to as simply “RSI”).  I can make a compelling case that the large amount of unnecessary mouse use required by ProE increases the risk of RSI, thereby making it unsafe.  I hope to convince you that it is in everyone's best interest for PTC to release an ergonomically optimized version of ProE as soon as possible. 

Since I cannot assume that you have had any personal experience with RSI, allow me to begin by quoting part of Wikipedia's entry on RSI:

“A repetitive strain injury (RSI), also called repetitive stress injury, cumulative trauma disorder or occupational overuse syndrome, is any of a loose group of conditions from overuse of the computer, guitar, knife or similar motion or tool. It is an occupational overuse syndrome affecting muscles, tendons and nerves in the arms and upper back; hence it is also known as work related upper limb disorder or WRULD. The medically accepted condition in which it occurs is when muscles in these areas are kept tense for very long periods of time, due to poor posture and/or repetitive motions.

It is most common among assembly line and computer workers. Good posture, ergonomics and limiting time in stressful working conditions can help prevent or halt the progress of the disorder. Stretches, strengthening exercises, massages and biofeedback training to reduce neck and shoulder muscle tension can help heal existing disorders.

Studies have related RSI and other upper extremity complaints with psychological and social factors. A large amount of psychological distress showed doubled risk of the reported pain, while job demands, poor support from colleagues, and work dissatisfaction also showed an increase in the report of pain, even after short term exposure.[1]”

What is this RSI-causing "overuse of the computer " mentioned in the above entry?  They are not talking about playing marathon-length iTunes playlists or watching 5 DVDs back-to-back.  "Computer overuse" or "computer use" is nothing other than the use of the mouse and keyboard, as directed by the Graphical User Interface, to operate software.

Blaming hardware for RSI misses the point entirely.  There is nothing intrinsic to the hardware itself that causes the mouse to be moved and clicked in an RSI-causing way.  The mouse is, in fact, the slave of the Graphical User Interface, which demands and defines the mouse’s repetitive use.  Software is the reason you move your mouse.  Software is the cause of RSI, not hardware.  The focus on hardware as an RSI source is like spraining your ankle running on uneven terrain, and then blaming the shoes.  RSI-causing static tension only enters the body because of the hand-eye coordination required by moving & clicking the cursor (hand) in the GUI (eyes).  The GUI controls the mouse.  It is the GUI that is responsible for the repetitive highly-controlled mouse movements, clicking, and long button depressions which lead to the buildup of RSI-causing static muscle tension.  No software = no mouse use = no RSI, regardless of any other risk factors. 

Through your design of the screen layout, you control the movement of the cursor on the screen, and hence the mouse use of the user. Since this mouse use is by definition the cause of mouse-related RSI, you the software designer become responsible for causing or contributing to mouse-related RSI in your users.  Software's control of the mouse hand identifies it as the ultimate cause of mouse-related RSI.  Once this simple but overlooked insight becomes established wisdom, the computing public will demand low-RSI-risk software, and punish those companies whose software remains high-risk.

So now that I've established the fact that software is the cause of RSI, where does that leave PTC?  Just for the fact of producing software, you are no worse off than before. GUI-induced RSI is a risk that society accepts in order to gain the usefulness of the computer, much like the risk of an accident entailed in driving a car.  Not everyone has an auto accident, and not everyone gets RSI.  Since the issue is now one of risk, however, the question becomes one of safety:  “Does the software pose an undue RSI risk?  In Low-risk software, all mouse use contributes directly to the desired result.  High-risk software is software that is full of unnecessary mouse use that does not add any information or value to the model, and only serves to increase the risk of RSI. 

The issue of liability may arise for software that is demonstrably and unnecessarily high-risk, and therefore unsafe.  Now let me show you how ProE has a demonstrable  and unnecessarily high risk of causing RSI.

ProE IS HIGH RSI-RISK SOFTWARE

One of the major reasons ProE is unsafe is because information and settings are unnecessarily concealed behind check boxes and scrolls in windows that are too small, when they should and could be visible and accessible to the user without extra mouse use.  This extra mouse use compounds the RSI-risk by increasing a user's general stress level due to their frustration with the unnecessary extra work. The presence of such concealed information exhibits a lack of reasonable care in creating a working environment where unnecessary mouse use, and therefore RSI-risk, is held to a minimum.  My colleagues and I estimate that ProE requires double the mouse use as necessary, which may even be a low estimate.

The following example illustrates what I am talking about.  To avoid detracting from the flow of my central argument, I have included further examples in the section “Why ProE is Unsafe”.

In the example above, in order to make a worst-case “view type” change, you need to

1.Click on check button to reveal view types – they are concealed behind the check box.

2.Move mouse to scroll and make a tiny, high- tension scroll movement since the choice in question is concealed at the end of the scroll window.

3.Move mouse and click on view type

If you had used additional radio buttons you would have avoided steps 1 & 2 entirely, which are therefore shown to be an unnecessary, avoidable, 200% increase in “computer use” that you are burdening your users with. 

Notice also how the scroll bars fill up most of the scroll space.  In the two scrolls in this window, there are only respectively one and two more choices available.  These few extra choices are concealed from the user behind a high-tension scroll movement.  This is terrible design!  Make the windows larger  to include all the options and get rid of the scroll bars!  ProE is replete with this particular design error.  Not only are the tiny scrolls a major source of static muscle tension in themselves, but the frustration of having to repeatedly perform unnecessary mouse use like this increases tension psychologically.

Having to click the “apply” button is a pain, as are all the “confirm” buttons throughout ProE.  Changes should be automatic, and confirmations “Yes” by default.  A more powerful, universal “undo” should then allow the opportunity to go back one step to where one was before.  This would save both mouse use and the user’s nerves. 

Making the whole window larger would eliminate or reduce having to click to switch between “Categories”.

All told, the unnecessary mouse work in ProE can add up to thousands of extra clicks and movements daily that can be demonstrated to have been avoidable with a more ergonomic interface, drastically increasing the risk of contracting RSI.  You might be held liable one day for not making these simple changes that would lower the RSI risk – not to mention make for more productive users.  Why have these simple, obvious things not already been done in the entire two decades of ProE's existence?

IN THE NOT-TOO-DISTANT FUTURE

Now imagine a user, John Doe, who after reading an article on how software contributes to RSI finally makes the connection between the pain in his body and the CAD software he has been using.  John's workplace has been ergonomically set-up, yet he has still gotten RSI while working on ProE full-time to meet a deadline.

He begins to reflect on all the unnecessary mouse use required by ProE. 

"Would I have gotten RSI to begin with," he reflects, "if the program had been designed better?"

The answer, in many cases, will be "No".

He can't help wondering how things would have turned out if PTC had cut their required mouse use in half, which would have been relatively easy for PTC to do, if they had only taken reasonable care in designing their product.  Recently he used new mouse tracking software to exactly determine the extra burden on his hand and arm due to unnecessary mouse use, with the following results:

John spent 82% of his total computer use hours using ProE, and 18% using various Microsoft products (including private use).  Of his ProE time, 40% of his mouse use could be categorized as avoidable (=unnecessary).  That meant that 33% of his total computer time was spent performing irrelevant mouse use, or, expressed another way, his computer use was 50% higher than it otherwise would have been.

The software could not only track the number of clicks and movements associated with unnecessary actions, but could also make an estimate of harmful, RSI-causing static muscle tension by analyzing the speed and movement characteristics of the mouse as the cursor approached a target.   This way an objective measure of RSI risk for any given software could be made.  The results of the “Tension-Tracker” test simply confirmed what he already knew.  ProE scored in the highest range of RSI-risk.

John can’t help but conclude that it was the extra 50% from this high-risk software that had pushed his body over the edge and sidelined him for weeks, putting his career as a ProE modeler in jeopardy.

Several of John’s colleagues had also developed RSI, so they get together and do some research.  From Wikipedia they learn about liability and negligence lawsuits:

“A products liability claim usually falls into one of three possible types:

1. •those claiming a design defect,
   

2. •those claiming a manufacturing defect, or
   

3. •those claiming a failure to warn.
   

A basic negligence claim consists of proof of

1.a duty owed,

2.a breach of that duty,

3.an injury, and

4.that the breach caused the plaintiff's injury.”

They reason that since ProE use constitutes computer use, and computer use is known to be a cause of RSI (as it has been in their case), that PTC owed it to their users to make the safest product possible, but had failed to exercise reasonable care in that regard, as evidenced by the results of John’s “avoidable mouse use” investigation.  Their injuries are medically verifiable.  Now all they need to do is to decide how they should approach the extra 50% of avoidable mouse use which they could demonstrate in court. 

“Is it a design defect, or a breach of the developer’s duty to make their software as ergonomic as possible?” they wonder.

Whatever the case, they note that the company had failed to issue an RSI-warning concerning the use of their software.

“So do we make a claim against the designer of ProEngineer, PTC, or against our employer, under the workman’s comp laws?”  They now ask themselves.  Whatever the case, they discover that other programs also have avoidable mouse use, so they expand the scope of their lawsuit to include them as well.  They also realize they are not the only ones with RSI, and use the reach of the internet to find others to join them in filing a class-action lawsuit.

In the end, though, they decide to go after their employer.  They win, he loses, and Liberty Mutual decides to raise their workman’s comp insurance rates for employers utilizing ProE.  They are starting to see this sort of thing too often.

RSI WILL IMPACT BOTTOM LINE

Is this scenario possible?  I believe it will not be long before the idea linking software to RSI-risk enters the public mind and the demand for low RSI-risk software becomes the norm.  In the future, it will not be necessary to prove that ProE caused a plaintiff’s RSI, but it will be enough to show that software in general significantly contributes to, or even causes RSI entirely, and that the software is unsafe (high RSI-risk).  Both of these things can be argued successfully about ProE, in my opinion, and given the litigious nature of American society, it is only a matter of time until it happens.

Once the mouse-tracking software has been developed to the extent envisioned in this scenario, you could find yourself being put in the position to explain why you continue to force a measurably high amount of extra, unnecessary mouse use on your users, when it could easily have been avoided. You are in the unique position  of being able to reduce the actual activity that leads to RSI.  A chair or mouse manufacturer can make an ergonomic product to make the best of a bad situation.  You can change the bad situation itself.

Unfortunately, my experience with PTC is such that I think that financial pressure, or the anticipation thereof, is the only thing that will really motivate you to  actually make the changes I am suggesting.  My PTC enhancement request detailing some of these changes,  (#4889848, October 2005, in the name of SysAdmin Kees Bouwman) was not answered by the time I left that company in mid-January 2006.  I checked in months later, and apparently it had still not been answered.

I also simply cannot believe that I am the only user, in the long history of the application, to suggest that these improvements be made.  Because nothing has changed, however, I assume that they are too low on your list of priorities for anything to ever be done about them.  Up until now, RSI complaints have been linked to hardware, and software users typically take what they are given, without imagining that they could demand change.  Instead of seeing the software as the cause of their RSI, they accept its shortcomings as a given.  My colleagues and I hereby demand change.

So, although I regret framing these arguments in terms of a lawsuit, by linking software to RSI (and thus RSI lawsuits), I am linking ProE's ergonomics to PTC's bottom line. This provides a strong financial incentive to make the changes that should have been implemented long ago.  Continuing to sell ProE without taking out all unnecessary mouse use is comparable to selling a car without seatbelts, airbags, or headrests, and with the same legal repercussions if there is an accident causing bodily harm.

WHAT DOES THE FUTURE HOLD?

Just as the growing awareness of safety has led to product design changes in a myriad of other industries, so too will the growing awareness of software's role in causing RSI lead to the demand for more ergonomic software. 

What kind of effect will this realization have on society?  I can foresee a number of possible developments:

RSI Lawsuits against software manufacturers who have demonstrably unsafe software.

Insurance rates (workman's comp) may be linked to software RSI risk.  Low risk = low rates, high risk = high rates.

It becomes a workplace safety issue subject to governmental regulation in Europe (and perhaps OSHA in the US)

Public outcry results in software manufacturers improving the ergonomics of their products! (Preferred result!)

I trust that you can now see how RSI is a software-associated medical safety issue, and that you must take up your responsibility for making your software as low-risk as possible.  I ask that you make a priority of this.  I am sure we would both agree that it is better for everyone for PTC to use their resources improving ProE rather than for defending themselves in court.

For the near term, I ask that you release an update as soon as possible that solves the information accessibility issue that I pointed out.  For the mid term, I request that you re-work the program interface to simplify function operations, and implement ergonomic guidelines throughout the program, such as the ability to access all functions / make all changes within 2-3 clicks.  Such ergonomic guidelines for software exist already, although I do not have a reference for you at the moment. You would do well to benchmark ProE against Catia V5 and SolidWorks, for they are better examples of usability & ergonomics. 

With just a little effort, you can make great usability improvements.  This will be good for your business as well as for the health and well-being of your users, and may save me the trouble of turning this cause into a website.  I'm right about this, and I intend to keep after PTC, in one way or another, until you make the changes.  I'm fighting for the health and professional future of myself,  my colleagues, and for all the silent RSI sufferers.  I look forward to hearing from you that PTC will release an ergonomically improved version of ProE in the near future.

Thank you for your time.

sincerely,

Gerrit Newton

We the undersigned fellow ProE users agree with this letter, and urge you to now exercise "reasonable care", and implement and distribute the requested changes as soon as possible. (“RSI” denotes users with some degree of RSI)