New hardware platform for Edge applications

At the SPS IPC Dri­ves 2018, Siemens showed a new­ly devel­oped hard­ware plat­form for edge appli­ca­tions as part of its Siemens Indus­tri­al Edge con­cept: The com­pact Simat­ic Edge Device works on the basis of the embed­ded indus­tri­al PC Simat­ic IPC227E, and fea­tures inte­grat­ed con­nec­tiv­i­ty to automa­tion on the machine lev­el. This allows man­u­fac­tur­ing data to be cap­tured and processed direct­ly at the point of pro­duc­tion. If there is a change to the frame­work con­di­tions under­ly­ing the indus­tri­al appli­ca­tion, Indus­tri­al Edge offers facil­i­ty to adjust soft­ware appli­ca­tions on the Edge Device, keep­ing them right up to date using func­tion­al, feed­back-free updates. The hard­ware comes with a closed all-met­al hous­ing, ensur­ing max­i­mum indus­tri­al func­tion­al­i­ty for flex­i­ble, main­te­nance-free use under even the harsh­est of con­di­tions. Rapid com­mis­sion­ing is guar­an­teed using pre-installed Edge software.

Industrial Edge

Siemens Indus­tri­al Edge affords users the oppor­tu­ni­ty to close the gap between clas­sic local data pro­cess­ing and cloud-based data pro­cess­ing to suit indi­vid­ual require­ments. Edge com­put­ing enables the local feed­back-free pro­cess­ing of large data vol­umes prac­ti­cal­ly in real time. Indus­tri­al Edge also enables users to cut the costs of data stor­age and trans­mis­sion, as large vol­umes of data are processed in advance and exclu­sive­ly rel­e­vant data is sub­se­quent­ly sent to the cloud or the com­pa­ny’s own inter­nal IT infra­struc­ture. Siemens Indus­tri­al Edge sup­ports cloud trans­mis­sion pro­to­cols for Mind­Sphere, the open cloud-based IoT oper­at­ing sys­tem from Siemens. In future, it will also sup­port Mes­sage Queu­ing Teleme­try Trans­port (MQTT), which will addi­tion­al­ly guar­an­tee the flex­i­ble exchange of data with oth­er sys­tems and clouds.

Savings through edge computing in the Siemens Electronics Factory Amberg

In the fac­to­ry in Amberg, a print­ed cir­cuit board cut­ting machine is used for the man­u­fac­ture of Simat­ic prod­ucts. Dur­ing a milling oper­a­tion, fine milling dust is pro­duced which exerts an aggres­sive impact on the machine. This can cause the spin­dle bear­ing to jam, and poten­tial­ly result in unsched­uled machine down­time. To pre­vent this from hap­pen­ing, a num­ber of the machine’s oper­at­ing para­me­ters are ana­lyzed using arti­fi­cial intel­li­gence with a view to detect­ing any anom­alies in spin­dle behav­ior which would indi­cate the pos­si­bil­i­ty of an impend­ing fail­ure. This is done by trans­mit­ting data picked up by the sen­sors to the Edge device for analy­sis. A machine learn­ing algo­rithm cal­cu­lates the anom­aly val­ue in real time. A rise above a pre­de­ter­mined thresh­old val­ue indi­cates that a machine fail­ure is immi­nent. This Edge appli­ca­tion enables cas­es of bear­ing ero­sion and machine down­time to be pre­dict­ed between 12 and 36 hours in advance of an actu­al fail­ure. In the event of anom­alies, the machine spin­dle can be exchanged as part of the next sched­uled ser­vice before a fail­ure can cause cost­ly unplanned downtime.