Why should you get involved in Standards Development?

Committee groupIt’s now been more than 30 years since I first learned about stan­dards, in the form of the Ontario Elec­tri­cal Code. I was study­ing elec­tric­i­ty and elec­tron­ics in high school, and Steve Struk, the Elec­tric­i­ty Teacher at Erindale Sec­ondary School in Mis­sis­sauga, intro­duced us to the rules. My first encounter with inter­na­tion­al stan­dards was 25 years ago, when I was tasked with doing some envi­ron­men­tal stress test­ing using a ther­mal, humid­i­ty and vibra­tion cham­ber at Ham­mond Man­u­fac­tur­ing in Guelph. Stan­dards were, and are, an impor­tant part of engi­neer­ing and tech­nol­o­gy, and they play increas­ing­ly impor­tant roles in busi­ness and occu­pa­tion­al health and safe­ty. Writ­ing stan­dards takes time. Lots of time. That time has to be pro­vid­ed by inter­est­ed peo­ple and orga­ni­za­tions who rec­og­nize the val­ue that stan­dards bring to their work and their busi­ness­es. Most peo­ple involved in stan­dards com­mit­tees are “paid vol­un­teers”, mean­ing that they are vol­un­teer­ing their time, but their employ­ers are pay­ing them for the time they spend engaged in stan­dards work. Some, like me, are true vol­un­teers, where the time spent on stan­dards work is giv­en with­out any com­pen­sa­tion except the knowl­edge that we are con­tribut­ing some small part to mak­ing Cana­da a bet­ter place to live and work.

So why should you get involved in stan­dards work? A recent SCC com­mu­niqué said it very well:

SCC’s mem­ber­ship in the Inter­na­tion­al Orga­ni­za­tion for Stan­dard­iza­tion (ISO) and the Inter­na­tion­al Elec­trotech­ni­cal Com­mis­sion (IEC) pro­vides Cana­di­an busi­ness, gov­ern­ment and con­sumer stake­hold­ers with the oppor­tu­ni­ty to sit at the table of glob­al experts mak­ing the rules that will then dic­tate the glob­al trade agen­da with­in their field. As a mem­ber of a com­mit­tee devel­op­ing an inter­na­tion­al stan­dard, Cana­da becomes part of a com­mu­ni­ty of nation­al experts cre­at­ing the stan­dard they need, and can ben­e­fit great­ly through greater involve­ment in the devel­op­ment process.

As a mem­ber on a tech­ni­cal com­mit­tee at ISO or IEC, a busi­ness can influ­ence the future of its mar­ket, get ear­ly access to cut­ting edge infor­ma­tion and define its com­pet­i­tive envi­ron­ment inter­na­tion­al­ly. Sim­ply put: par­tic­i­pa­tion in inter­na­tion­al stan­dards devel­op­ment pro­vides Cana­di­ans with an oppor­tu­ni­ty to influ­ence a doc­u­ment that could affect their lives or busi­ness in the future.”

Get­ting involved in stan­dards writ­ing is reward­ing, chal­leng­ing, work. Get­ting involved gives you a chance to con­tribute your knowl­edge and exper­tise to Canada’s future, and pro­vides an “…oppor­tu­ni­ty to influ­ence a doc­u­ment that could affect their lives or busi­ness in the future.”

Get involved. Con­tribute. It’s worth it!

Digiprove sealCopy­right secured by Digiprove © 2014
Acknowl­edge­ments: Stan­dards Coun­cil of Cana­da
Some Rights Reserved

Risk Assessment Blunders

Analysis TeamUpdat­ed 7-Jul-2014

Recent­ly I read a blog post writ­ten by David Cant, called, “Are You Mak­ing These Risk Assess­ment Blun­ders?”. Writ­ing in the UK, Mr. Cant spoke to some of the com­mon kinds of prob­lems that can occur when employ­ers con­duct risk assess­ments. Many his points are equal­ly applic­a­ble to machine build­ing:

  1. No Risk Assess­ment — This seems self evi­dent, since you can’t work to con­trol what you don’t know exists. Unfor­tu­nate­ly, some lia­bil­i­ty lawyers have advised clients NOT to con­duct risk assess­ments, on the basis that you can’t be blamed for some­thing you knew noth­ing about. This posi­tion is chang­ing, since risk assess­ment has become a fun­da­men­tal piece of the machin­ery design process and is includ­ed in US, Cana­di­an, Inter­na­tion­al, EU, and Aus­tralian stan­dards to name a few. What remains as a lia­bil­i­ty are poor­ly done risk assess­ments, and those can cer­tain­ly hurt a com­pa­ny in a lia­bil­i­ty case. See Point 3.
  2. Not Prop­er­ly Iden­ti­fy­ing Haz­ards — This is anoth­er big one. Machine builders some­times fail to under­stand the haz­ards that are incor­po­rat­ed into their machin­ery, espe­cial­ly in the case of sys­tem inte­gra­tors. Iden­ti­fi­ca­tion is the first step, analy­sis is the sec­ond step. Once the haz­ard has been iden­ti­fied, the asses­sors must ana­lyze the haz­ard to under­stand the mag­ni­tude of injury that can occur, e.g., a paper cut, hear­ing loss, per­ma­nent mus­cu­loskele­tal dis­or­der, fatal­i­ty, etc.
  3. Cre­at­ing an inad­e­quate risk assess­ment — If you “phone it in”, any knowl­edgable per­son will be able to see that, and you can be sure that pros­e­cu­tors will bring that to the atten­tion of the court! Any claim that is made will be refut­ed by a knowl­edgable per­son hired by the pros­e­cu­tion.
  4. Assum­ing Safe­ty Because You Have a Doc­u­ment — A risk assess­ment report nev­er pro­tect­ed any­one from harm. Actions based on the report are what pro­tect peo­ple from harm. Sim­ply com­plet­ing a risk assess­ment and stick­ing the report in the design file fix­es noth­ing, and going back to the lia­bil­i­ty dis­cus­sion, will cre­ate prob­lems when it becomes clear that a) You knew about the risks, b) You thought out mit­i­ga­tion mea­sures for the risks, but c) You failed to act on your own rec­om­men­da­tions.
  5. Not involv­ing your employ­ees — I would change this to read, “Not involv­ing the cus­tomer or end user.” Get­ting a risk assess­ment done by an out­side indi­vid­ual is like­ly going to result in a poor out­come, if that per­son does not work with the knowl­edgable peo­ple in your orga­ni­za­tion. If this is an inter­nal risk assess­ment for a process used in-house, then you need the work­ers involved with the process to be part of the risk assess­ment team. You also need the process design­ers (elec­tri­cal, mechan­i­cal, soft­ware, etc.). If this is a prod­uct devel­op­ment risk assess­ment, then you will need prod­uct design­ers, mar­ket­ing or sales peo­ple, and end-users. In all cas­es you may also need out­side experts to help with haz­ard analy­sis and mit­i­ga­tion, since you may not have the exper­tise in-house. A good exam­ple of this is the use of indus­tri­al lasers into process­es.
  6. Assess­ing risk gener­i­cal­ly — this goes back to point 3 — a risk assess­ment has to be rel­e­vant to the process, prod­uct or ser­vice that it describes. Using a risk assess­ment devel­oped by some­one else for some oth­er sim­i­lar prod­uct, process or ser­vice lim­its your think­ing to just what they decid­ed was rel­e­vant. Bet­ter to start with a blank sheet, and only at the end look at other’s work to see if you may have missed any­thing they includ­ed.
  7. Doing the risk assess­ment after the machine has been designed, built and put into use — This is a com­mon prob­lem, and espe­cial­ly affects work­places that buy off-the-shelf machin­ery and used machin­ery. Many machine builders are not famil­iar with risk assess­ment or the ben­e­fits to their prod­ucts, their lia­bil­i­ty expo­sure, and their rep­u­ta­tion in the mar­ket­place, let alone the ben­e­fits to their cus­tomers. Con­duct­ing the risk assess­ment after the equip­ment is in use means that the first stage of the Hier­ar­chy of Con­trols can­not be used. Since this is the only stage that can achieve 100% risk reduc­tion, this is a HUGE loss. In addi­tion, this means that any changes need­ed to mit­i­gate risk require expen­sive changes when the machin­ery is already in pro­duc­tion, com­pound­ing the expense with lost pro­duc­tion. If you have no oth­er option — for instance, you’ve tak­en over an exist­ing busi­ness and no risk assess­ments have been done in the past, then late is bet­ter than nev­er, but it should be a last resort and not the first choice. Spe­cial thanks to Dou­glas Flo­rence for sug­gest­ing this added blun­der!

All of this brings me to some impor­tant stan­dards. The Cana­di­an Stan­dards Asso­ci­a­tion released CSA Z1002, Occu­pa­tion­al health and safe­ty — Haz­ard iden­ti­fi­ca­tion and elim­i­na­tion and risk assess­ment and con­trol, in 2012. This is a land­mark stan­dard, because it address­es work­place risk assess­ment. No oth­er stan­dards devel­op­ment orga­ni­za­tion has released any­thing quite like it. There are some inno­v­a­tive ideas in the stan­dard, includ­ing the idea of “risk trans­fer.” This con­cept explains how risk is trans­ferred from the devel­op­er or sup­pli­er of a prod­uct, process or ser­vice, called the “exter­nal con­text”, to the user orga­ni­za­tion, called the “inter­nal con­text”, and then to the work­er. At each trans­fer point, the risk should have been reduced so that the worker’s expo­sure is as low as pos­si­ble, and cer­tain­ly no high­er than per­mit­ted by law.

In the machin­ery world, the “moth­er stan­dard” is ISO 12100, Safe­ty of machin­ery — Gen­er­al prin­ci­ples for design — Risk assess­ment and risk reduc­tion. This stan­dard affects many of the machin­ery safe­ty stan­dards devel­oped world­wide, includ­ing CSA Z432 and the ANSI B11 stan­dards, and is har­mo­nized under the Machin­ery Direc­tive as EN ISO 12100. This stan­dard lays out the process for safe machine design and pro­vides the frame­work for machin­ery risk assess­ment. A com­pan­ion doc­u­ment, ISO/TR 14121–2, Safe­ty of machin­ery — Risk assess­ment — Part 2: Prac­ti­cal guid­ance and exam­ples of meth­ods, pro­vides guid­ance on how to con­duct a risk assess­ment, and offers up some exam­ple tools that could be used for this pur­pose. In the US, ANSI pub­lish­es ANSI B11-TR3, Risk Assess­ment and Risk Reduc­tion — A Guide to Esti­mate, Eval­u­ate and Reduce Risks Asso­ci­at­ed with Machine Tools, pro­vid­ing solar guides to users of the B11 fam­i­ly of stan­dards.

Need to know more? Con­tact me for infor­ma­tion train­ing and risk assess­ment work in your facil­i­ty!


Why YOU need a Product Safety Certification Strategy that Works!

How compliance efforts go wrong

Clients call me when they have prob­lems. They sent some prod­uct to a cer­ti­fi­ca­tion body, some tests were done, and the prod­uct failed. Now what? Usu­al­ly, the deliv­ery date for the cer­ti­fied prod­uct is approach­ing quick­ly, and no time is avail­able to react. Deliv­ery is delayed, and the cus­tomer is upset. The busi­ness may even be lost.

This process goes wrong for a num­ber of rea­sons:

  1. The end goal for the cer­ti­fi­ca­tion and the inter­me­di­ate require­ments were not con­sid­ered
  2. The cer­ti­fi­ca­tion body chose the stan­dard, and the man­u­fac­tur­er had no idea what the stan­dard required
  3. The sam­ples were not care­ful­ly pre­pared for the cer­ti­fi­ca­tion body
  4. Infor­ma­tion was miss­ing or par­tial­ly miss­ing

Certification Motivation

What moti­vates a man­u­fac­tur­er to cer­ti­fy a prod­uct? The deci­sion to cer­ti­fy comes about for a vari­ety of rea­sons, includ­ing:

  1. Legal require­ment for the mar­ket
  2. Cus­tomer request
  3. Lia­bil­i­ty lim­i­ta­tion
  4. Mar­ket­ing advan­tage over competitor’s offer­ings

I’ve talked about the dif­fer­ences between CE Mark­ing and tra­di­tion­al Cer­ti­fi­ca­tion process­es in a few arti­cles, includ­ing this one recent­ly, so I won’t repeat myself here. In Cana­da, Fed­er­al [1, 125(m)(iii)] and Provin­cial law [2, 113 (1) ©] require any­one sell­ing an elec­tri­cal prod­uct to ensure that it bears an elec­tri­cal safe­ty mark, and this is sup­port­ed in the Cana­di­an Elec­tri­cal Code [3, 2–024]. This require­ment is often over­looked in the industrial/commercial mar­ket where post-instal­la­tion equip­ment inspec­tion, called “Field Eval­u­a­tion”, is typ­i­cal.

In the US, each State has a slight­ly dif­fer­ent approach, so you will need to check out the require­ments in the states where your prod­uct is sold so that you can ensure com­pli­ance with the local require­ments. From a work­place per­spec­tive, the US OSHA requires that all elec­tri­cal prod­ucts used in the work­place bear a mark from a Nation­al­ly Rec­og­nized Test­ing Lab­o­ra­to­ry (NRTL) [5].

Cer­ti­fy­ing your prod­uct pro­vides some degree of lia­bil­i­ty lim­i­ta­tion, in that it shows that you met the min­i­mum legal require­ments for your mar­ket­place. To lim­it your lia­bil­i­ty effec­tive­ly, you will need to do more than just meet the min­i­mum require­ments, and you will need doc­u­men­ta­tion of every­thing done to meet or exceed those require­ments.

If you offer cer­ti­fied prod­ucts in a mar­ket where this is uncom­mon, you have a mar­ket­ing advan­tage as long as your cus­tomers under­stand the ben­e­fits cer­ti­fied prod­ucts bring. If cer­ti­fi­ca­tion is rare in your mar­kets, you may need to under­take some edu­ca­tion­al mar­ket­ing to help cus­tomers “get it”.

Anoth­er key point needs to be con­sid­ered: Prod­uct vol­ume. Cer­ti­fi­ca­tion costs mon­ey and takes time. If you are sell­ing less than 200 units per year of a giv­en prod­uct in your line, cer­ti­fi­ca­tion for that prod­uct is unlike­ly to be fea­si­ble. For prod­uct vol­umes from one to 200 units per year, Field Eval­u­a­tion pro­vides a much more time and cost-effec­tive way to get your prod­uct marked.

The dia­gram shows the gen­er­al process flow for this activ­i­ty. If you are choos­ing to use Field Eval­u­a­tion instead of Cer­ti­fi­ca­tion, sub­sti­tute “Field Eval­u­a­tion” wher­ev­er you see “Cer­ti­fi­ca­tion” in the dia­gram.

Flow chart showing certification process flow.
Cer­ti­fi­ca­tion Process Flow

Selecting standards

Select­ing the “right” stan­dard for your prod­uct can be a chal­lenge, espe­cial­ly in the indus­tri­al mar­ket where prod­ucts are often high­ly spe­cial­ized, “one-off” prod­ucts. In many of these cas­es, no stan­dard exists that specif­i­cal­ly cov­ers the prod­uct. For rel­a­tive­ly sim­ple prod­ucts, or for prod­ucts that are very com­mon, like TVs, com­put­ers, and audio-video equip­ment, there are “Prod­uct Fam­i­ly” stan­dards that specif­i­cal­ly cov­er these types prod­ucts.

Not every stan­dard is a cer­ti­fi­ca­tion stan­dard. Most of the cer­ti­fi­ca­tion stan­dards are focused on elec­tri­cal and fire safe­ty. The con­cerns are the pre­ven­tion of elec­tri­cal shock, arc flash, and fire. Cer­ti­fi­ca­tion stan­dards will typ­i­cal­ly include spe­cif­ic tests that must be passed to show com­pli­ance with the require­ments. Design stan­dards, on the oth­er hand, will pro­vide gen­er­al per­for­mance require­ments and some­times pre­scrip­tive fea­ture require­ments, but no test require­ments. This is typ­i­cal in the indus­tri­al machin­ery sec­tor where stan­dards like CSA Z432 [6] and the ANSI B11 fam­i­ly [7] of stan­dards apply. In these cas­es, you may be able to have the prod­uct cer­ti­fied for elec­tri­cal safe­ty, but not for machin­ery safe­ty. This does not elim­i­nate cor­po­rate lia­bil­i­ty for the machin­ery haz­ards, requir­ing man­u­fac­tur­ers to be knowl­edge­able and dili­gent in apply­ing design stan­dards.

Developing a Certification Strategy

To devel­op a sound strat­e­gy, I rec­om­mend a “bot­tom-up” approach. To apply this idea, start with the bill of mate­ri­als for the prod­uct. Look first at the pur­chased prod­ucts: How many of these items are either already cer­ti­fied by their man­u­fac­tur­er? All of the cer­ti­fied items can be elim­i­nat­ed from fur­ther con­sid­er­a­tion for the moment. Next, con­sid­er the pur­chased but un-cer­ti­fied prod­ucts. Con­tact all of your sup­pli­ers to deter­mine which of these prod­ucts can be pur­chased cer­ti­fied, and adjust the bill of mate­ri­als to reflect the part num­bers for the cer­ti­fied ver­sions.

Now, the hard­er part. All of the remain­ing items on the bill of mate­ri­als need to be looked at for cer­ti­fi­ca­tion. Any­thing that can­not or need not be cer­ti­fied, e.g., nuts and bolts, oth­er mechan­i­cal parts that are not pres­sure bear­ing, etcetera, can be exclud­ed from con­sid­er­a­tion. You now have a short list of uncer­ti­fied com­po­nents that require cer­ti­fi­ca­tion.

For each item on the short list, research the stan­dards avail­able. The Scope of the stan­dards will help guide you regard­ing their applic­a­bil­i­ty. Once you have a matched list of com­po­nents and stan­dards, you can extend that research to include the top lev­el prod­uct.

Now you have the begin­ning of a com­pli­ance strat­e­gy. The next piece of the puz­zle involves the inter­nal eval­u­a­tion of each com­po­nent against the stan­dards cho­sen. This gives you the abil­i­ty to revise your think­ing, either of the stan­dards you chose or of the design and con­struc­tion of the com­po­nent. Mak­ing good choic­es at this stage to either cor­rect issues found in the design or con­struc­tion of the com­po­nent, or in the selec­tion of the stan­dard, can save you huge amounts of time and effort once the cer­ti­fi­ca­tion body gets involved.

Once the com­po­nents have suc­cess­ful­ly passed the inter­nal “pre-com­pli­ance” eval­u­a­tion, you can get the cer­ti­fi­ca­tion body involved, and start the for­mal com­pli­ance process for each com­po­nent. As this part of the process pro­gress­es, the cer­ti­fi­ca­tion body may have addi­tion­al ques­tions or requests for infor­ma­tion. To reduce these in-process ques­tions, make sure that each com­po­nent is clear­ly iden­ti­fied, that you have unique part num­bers for each part and that you have pro­vid­ed infor­ma­tion on the mate­ri­als used in the con­struc­tion of the com­po­nent, as well as detailed engi­neer­ing draw­ings.

As the com­po­nent cer­ti­fi­ca­tion work pro­gress­es, you can start on the top lev­el prod­uct cer­ti­fi­ca­tion work. The top-lev­el prod­uct needs to go through the same sort of inter­nal pre-com­pli­ance process as the com­po­nents so that you can be as cer­tain as pos­si­ble that the prod­uct will meet the require­ments when it gets to the cer­ti­fi­ca­tion lab.

Prepa­ra­tion of the data pack­age and the sample(s) of the top-lev­el prod­uct that will be sub­mit­ted must be done care­ful­ly. Con­struc­tion of the sam­ples must match the man­u­fac­tur­ing draw­ings and instruc­tions as close­ly as pos­si­ble. Once every­thing is ready, the sam­ples can be sub­mit­ted for eval­u­a­tion.

Working with a Certification Body

Deal­ing with a Cer­ti­fi­ca­tion Body can be very chal­leng­ing. Much of the expe­ri­ence will be based on the project engi­neer that is respon­si­ble for your product’s eval­u­a­tion. It’s impor­tant to set up a good rela­tion­ship with this per­son at the begin­ning because once prob­lems start to crop up in the lab, you will need to be able to talk to this per­son. Mak­ing sure that you have the “right” stan­dards select­ed for your prod­uct is essen­tial, and the project engi­neer must agree with you. They can refuse to cer­ti­fy a prod­uct if they feel that the stan­dard cho­sen is incor­rect, and since they have the final word, there is no argu­ing with them. An open dis­cus­sion at the begin­ning of the project to dis­cuss the stan­dards select­ed is an excel­lent place to start. If your ideas and theirs devi­ate in a big way, you may have to com­pro­mise on their selec­tion, or worse, stop the project and review the prob­lems encoun­tered.

Once the prod­uct is cer­ti­fied, the Cer­ti­fi­ca­tion Body will con­duct reg­u­lar audits on the man­u­fac­tur­ing facility(ies) to make sure that the pro­duc­tion test­ing is being done, pro­duc­tion records are kept, and that the QA pro­grams are ensur­ing that only good prod­uct leaves the plant.

An impor­tant part of the QA process is the Cus­tomer Com­plaints Pro­gram. Man­u­fac­tur­ers must have a pro­gram in place to record cus­tomer com­plaints and to respond to those com­plaints. A deci­sion tree that helps cus­tomer ser­vice rep­re­sen­ta­tives dif­fer­en­ti­ate between safe­ty-relat­ed and non-safe­ty relat­ed com­plaints should be devel­oped. Safe­ty-relat­ed com­plaints should result in engi­neer­ing review of the prob­lems and deter­mi­na­tions about the cause of the prob­lems. If these are relat­ed to man­u­fac­tur­ing or design issues, and espe­cial­ly if these are relat­ed to com­pli­ance with the require­ments of the cer­ti­fi­ca­tion stan­dard, a recall of the prod­uct may be need­ed. If this is the case, get the Cer­ti­fi­er involved as soon as pos­si­ble. Fail­ure to act, and fail­ure to inform the cer­ti­fi­er can result in the cer­ti­fi­ca­tion being revoked.

References

[1]     Cana­da Labour Code, [online]. Avail­able: http://laws-lois.justice.gc.ca/eng/acts/L-2/. Accessed: 2018-01-10.

[2]     Ontario Elec­tric­i­ty Act — Mark­ing require­ments Avail­able: https://www.ontario.ca/laws/statute/98e15?search=electrical+safety#BK245. Accessed: 2018-01-10.

[3]     Cana­di­an Elec­tri­cal Code, CSA C22.1. 2012.

[4]     Nation­al Elec­tri­cal Code, NFPA 70. 2014.

[5]     Occu­pa­tion­al Safe­ty and Health Stan­dards, 1910 Sub­part S, Elec­tri­cal, Instal­la­tion and Use. 29 CFR 1910.303(b)(2), [online]. Avail­able: https://www.osha.gov/pls/oshaweb/owadisp.show_document?p_table=STANDARDS&p_id=9880. Accessed: 2014-01-27.

[6]     Safe­guard­ing of Machin­ery. CSA Z432. 2016.

[7]     Safe­ty of Machines. B11.org, [online]. Avail­able: http://b11standards.org/standards/. Accessed: 2014-01-27.