Avoiding costly machine failure

As man­u­fac­tur­ers con­tin­ue to auto­mate their fac­to­ries and con­nect them with intel­li­gent sen­sors, the data col­lect­ed is pro­vid­ing crit­i­cal infor­ma­tion on the health and remain­ing life of their machin­ery, enabling scal­able pre­dic­tive main­te­nance.

The idea of under­stand­ing the health of machin­ery, oth­er­wise known as con­di­tion-mon­i­tor­ing, is noth­ing new. As a prin­ci­ple, it’s been around for gen­er­a­tions, but online tech­nol­o­gy saw wide­spread adop­tion in the 1990s in the aero­space indus­try.

Unfor­tu­nate­ly, per­form­ing it at indus­tri­al scale has remained a pipedream for every­one oth­er than the priv­i­leged few who have been able to afford the sky-high costs of scal­ing it up, and employ­ing an army of expen­sive mon­i­tor­ing and data experts to extract the crit­i­cal insights.

Prog­nos­tics goes beyond con­di­tion-mon­i­tor­ing by show­ing the remain­ing use­ful life of machin­ery. As man­u­fac­tur­ers delve fur­ther into Indus­try 4.0, they are increas­ing­ly seek­ing these tech­niques to reduce cost­ly down­time. For­tu­nate­ly, the democ­ra­ti­sa­tion of enabling tech­nolo­gies, such as cloud com­put­ing and the inter­net of things (IoT), is now set to accel­er­ate these from their lim­it­ed-scal­a­bil­i­ty phas­es.

“Prog­nos­tics soft­ware takes in data from Indus­try 4.0 machin­ery and auto­mat­i­cal­ly builds a pic­ture of machine health as well as the remain­ing use­ful life,” says Alex Hill, chief com­mer­cial offi­cer at Sens­eye, a young UK com­pa­ny whose machine-learn­ing-based soft­ware for pre­dic­tive main­te­nance already helps sev­er­al For­tune 100 com­pa­nies pre­vent down­time from machine fail­ure on their pro­duc­tion lines.

“By automat­ing con­di­tion-mon­i­tor­ing and prog­nos­tics analy­sis, you can do pre­dic­tive main­te­nance at scale, so you know which machines are cur­rent­ly healthy and which aren’t, and you know which ones will be healthy or not in a giv­en time­frame.”

Sens­eye helps man­u­fac­tur­ers avoid down­time and save mon­ey by auto­mat­i­cal­ly fore­cast­ing machine fail­ure. Its unique machine-learn­ing algo­rithms allow it to be used on any machine from any man­u­fac­tur­er, tak­ing infor­ma­tion from exist­ing indus­tri­al IoT sen­sors and plat­forms to diag­nose fail­ures auto­mat­i­cal­ly and pro­vide the remain­ing use­ful life of machin­ery.

If unplanned down­time hap­pens, for exam­ple in the auto­mo­tive indus­try, every minute costs £50,000, Mr Hill says. “It’s an incred­i­ble amount of mon­ey that our cus­tomers want to avoid los­ing; if they have a down­time event they can nev­er make that back. By using our tech­nol­o­gy, indus­tri­al com­pa­nies have reduced down­time by 50 per cent in three months. It’s incred­i­bly quick to see results.”

Cen­tral to Senseye’s approach is cloud-based tech­nol­o­gy that doesn’t require experts in con­di­tion-mon­i­tor­ing or data sci­ence to oper­ate. It’s designed to be used by the main­te­nance team, not IT staff.

Sens­eye helps man­u­fac­tur­ers avoid down­time and save mon­ey by auto­mat­i­cal­ly fore­cast­ing machine fail­ure

Pre­vi­ous­ly, pre­dic­tive main­te­nance was so dif­fi­cult and expen­sive to do at scale because of the need for these skills. By not rely­ing on human exper­tise to analyse the data, man­u­fac­tur­ers can now enjoy pre­dic­tive main­te­nance at scale, going from a few machines to a few thou­sand with­out requir­ing
more resources.

“We do all the dif­fi­cult analy­sis work for them and they just get infor­ma­tion about what machines are caus­ing prob­lems and how long they’re like­ly to live,” says Mr Hill. “We con­nect with what­ev­er data source is already being used so instal­la­tion can take any­thing from half an hour to two weeks, depend­ing on what’s in place, mak­ing the whole set-up process as pain­less as pos­si­ble.”

In the com­ing years, pre­dic­tive main­te­nance is not only poised to reduce fac­to­ry down­time dras­ti­cal­ly, but it will also enable servi­ti­sa­tion in indus­tri­al sec­tors. The con­cept has already trans­formed the aero­space indus­try by chang­ing the busi­ness mod­el of many air­craft man­u­fac­tur­ers, for exam­ple, from phys­i­cal­ly offload­ing assets to sell­ing and deliv­er­ing flight hours.

The evo­lu­tion of servi­ti­sa­tion will mean indus­tri­al man­u­fac­tur­ers will begin to sell machine capa­bil­i­ty rather than machines them­selves. For exam­ple, instead of the man­u­fac­tur­er sell­ing a weld­ing robot, it will sell a num­ber of welds per hour, mov­ing from a mod­el of sell­ing hard­ware to offer­ing prod­ucts as a ser­vice.

“Deliv­er­ing a prod­uct as a ser­vice requires a high degree of auto­mat­ed data analy­sis when you’re look­ing at the indus­tri­al scale of hun­dreds or tens of thou­sands of machines,” says Mr Hill. “Pre­dic­tive main­te­nance, with­out large deploy­ment or ongo­ing costs and with a return on invest­ment of months, now real­ly is achiev­able.”

For more infor­ma­tion please vis­it senseye.io