Even though I agree in part, personally I would have to respectfully disagree.
- Professional indemnity and public liability.
I have also found it very difficult to find free information on this subject. I have been tasked with writing a risk assessment for a robot built in 1999. Have been quoted $9,000 for an assessment from a well known safety company. The RIA has some publications available to purchase for this project.
The publication we are looking at purchasing is called 360 Robot super safety bundle.
There are many standards referenced, and the means to create an assessment are included.
documentation on safety standards is not free and education on topic is rather limited and kind of obscure. available courses are expensive and few.
the best way to start is by getting involved in safety review - pretty much anyone can be involved but translating evaluated safety level into functional circuit is not like that. one need to get familiar with what the safety devices are and how to avoid pitfalls that would ruin or diminish the actually achieved safety level.
there are many ways one can get trapped into something that will do just that. this is why safety review should include multiple people from different disciplines.
i was creating so many machines for years (not just robotic cells). all of them needed PHSR (pre-health and safety review). one thing that i found curious is that most of safety guys doing PHSR have background in mechanical engineering. this means that they will have expert knowledge of various hazards and forces (shear, pinch etc). however modern machines rely more and more on electronic guarding and many of the professionals that are responsible for safety do not understand details and intricacies of a safety circuit. many of them cannot tell if the actual implementation is correct or not.
it is not rare that they will simply observe cause and effect without knowing if particular function is accomplished with required safety level or not. some have teamed up with another PENG that check circuits and software. they would collect drawings and lot of photos to show to that other person.
several have asked to explain what the circuit does and how it meets the standards. those usually have thanked me for pointing out things that they did not notice or consider or know about. but some would not ask much, they would just ask for files, do entire evaluation on their own and be on their way.
to see if they really pay attention i would intentionally change a thing or two and look if they will spot it. it turns out that 99% of time they did not catch a thing. so i would insist on going back over wiring or software together with something silly like "technician is new and tends to make mistake, better double check that", then i would "find it" and explain it.
example 1
traditional safety circuits were hardwired. but one little detail such as lack of monitoring ruins safety level although circuit appears to function perfectly.
example 2
practically all modern safety circuits are using some form of a safety controller (safety PLC). depending on product type it is possible to mix safe and non-safe signals (or safe signals but of a lesser integrity). someone not familiar with it will not be able to understand if result is proper or not. and often solution is deemed safe because cell "includes safety PLC, safety sensors etc." even if logic is complete mess and uses things that are not allowed.
example 3
some people try to use own interpretation of what safety devices do and how they detect fault to try to come up with their own version of circuits by replacing safety rated components with not safety rated ones. years ago i actually had to step in and warn against such practices as one determined user kept on posting own "articles" and "instructions" on doing just that on one of the forums. such things are great for an academic discussion and review (so it can be shown where the thought process failed). but it is completely different story (inappropriate and not safe) to share and publicize ill conceived ideas specially when they put life and limb at risk.
many clients simply admitting that they don't know enough about it and require highest safety level on everything, no matter the risks or complexity of machine. but as mentioned before, integrator too may not know what they are doing, they will buy required components and put them together. then the PHSR guy will not quite understand what he is looking in and if the implementation is correct. in other words there is a lot of things out there that should not be approved. both integrators and safety engineers are doing that for living but they are humans too and every now and then they manage to get things wrong. so better use your own head and do not assume all is good because there is a label saying it was approved.
so what is the solution?
proper education is the best way to get things done since more and more products are introduced and implementations are different. personally, i would be in favor of periodic mandatory competency evaluation where candidates need to pass maze of pitfalls including all commonly used technology - new and old.
Something I stumbled across which a little out of date now and directed towards US market, but some good principled reading here.
Microsoft PowerPoint - RIA Safety Cct Design 13-09-30.pptx (robotics.org)
A discussion post conveys zero lawful risk and 1000 individuals attempting to settle similar issues will think of innumerable choices, and some will miss subtleties.
What I disdain most, If I'm attempting to fix a 1976 clothes washer, there is a youtube video for precisely that.. Attempting to plan a protected working envelope and so forth; and you don't track down anything
and we are having one right now, aren't we?
btw your example is comparing apples and oranges.
fixing a 1976 clothes washer deals with making that cloth washer operational, there are plenty of videos and discussion on internet how to make robot operational as well and - this is one of top places for that.
but i could find no video on the youtube showing how to make 1976 clothes washer safer by doing risk analysis and then implementing safety circuit. probably because functional cloth washer is not inherently dangerous and - functional robot arm is. which is also a reason that there is no safety standard (in 1976 or today) requiring things like control reliable redundant safety circuit with cross channel monitoring on a washing machine.
but if you want to see what the process may look like, let's see what the potential hazards involving 1976 clothes washer could be:
1. chocking
2, drowning
3. poisoning
4. burning
5. cutting/pinching/shearing
6. crushing
7. electrocution
i don't think there are people maimed or dead because of first six. washer is too big to be swallowed, too small to be drowned in, risk of poisoning may be real by swallowing detergents or soapy water but the risk here pertains to safety of the detergents and home plumbing rather than safety of washing machine. getting burned by washer is possible but difficult and degree of burn would be small. getting crushed by washer is possible but if you are strong enough to tip it over, it is extremely unlikely to result in any serious damage. and you do not have to place washer on top shelf of your bookcase or duct-tape it to the wall or balance it on top of four broomsticks, you may place it on the ground level where it is meant to be.
so the only reasonable risk is that of electrocution and that too is easily remedied by proper grounding. and that is easy to do without any special training or skill since washer is all metal construction so connecting one ground wire is not exactly rocket science.
in comparison to that, safeguarding of the robot is a much more complex subject. i am not a legal expert but advising on how to implement robot safety is probably not much different than offering medical advice. both deal with safety and consequences of doing it wrong could lead to serious harm and death. and i am pretty sure that is something that may have some sort of liability attached - even if a new forum member in his first time post may suggest otherwise. 
Is that an OEM instructional video to repair the washing machine, or provided by monkey_ballz76?
I wonder why Dr Sandra Lee has legal disclaimers at the start of her youtube videos?
Public Forums should be treated in the same way IMHO as I believe the copy and pasting and quick fix solution hunters out there do not always fully appreciate the foundations of this information before applying it and then complain when it doesn't work as they cannot solve it for themselves.
But will claim credit for it if it does work.
You see this when providing code examples sometimes, so why should safety be any different.
Then could extend to include statements of how to bypass safety devices.....and I'm not going there.
Totally agree with panic mode, apples and oranges.
I am not a legal expert either and I stand by my initial statement.
Pushing that aside though, a safety section could potentially cause harm, injury or fatality.
For that reason, it should be risk assessed.
Not by us, but by the owner of Robot Forum Werner Hampel.
I can see many potential hazards which would prevent me from being involved in a safety section.
if there is one thing clear for anyone that visits forum like this regularly, it is that a lot of forum members, most are consumers but very few are actually providing content that addresses questions.
that is because vast majority of members do not want to invest time and money into anything - specially in training, reading manuals or even asking questions in clear and concise manner. in other words many will happily take shortcuts where ever they please.
and just like with Drake equation... very few people will bother to research topic, even fewer will understand what they have read, and much much fewer will know what standards to follow to comply with rules in their own country, not to mention reading them and understanding them....
if that is not a definition of dangerous i don't know what it is.
those that are serious about PHSR of an industrial cell will invest time, money and effort to do it right or hire professionals. just watching 10-20min YouTube video simply does not cut it.
As always, right on the money.
These are those potentials I've been referring about.
This thread has just reminded me of the following thread which I'll add a post to:
It is unresolved as yet, but is what I would consider a perfect example of how this forum should approach safety related issues and kudos for EPeters1 for their attitude towards the information provided and kudos for byrol too.
