MechChem Africa January 2019

⎪ Computer-aided engineering ⎪

Data differentiates All areas of improvement dependon thevalue of thedata collected.Whydomodernday cars almost never break down? “Because the data collected over the years and in real time on the car itself enables design improvements and accurate problem prediction. “Data is the differentiator: starting with collection and visualisation and then using analytics to link the cyber and the physi- cal worlds. This can now be used to write automated self-optimising command and control routines. Humans no longer need to

be involved in controlling every step. Machines can take care of the standard stuff,” suggests Teifel. In a further step, he says that further analysis can be used to control and learn from the past – and forthefuture.Thisisabout Artificial Intelligence (AI) and machine learning. In addition, data from a host of similar events can be collatedsoastolearnfrom them too, which is the ultimate goal of machine learning and AI – “and the fastest to implement will be the winners”. From an implementa- tion point of view, he cites six important dimensions to look at: the industry/ business model; the busi- ness focus for the roll-out; the technologies involved; the ITsolutions; the imple- mentation strategy; and the change management solution. “The 4IRmodel chosen for every company will be different,” he says. “A

Thulani Mazibuko demonstrates the use of PTC Creo’s new topology optimisation capability for a robot gripper at the Creo 5.0 Launch event at Swartkops Raceway. Inset 1: gripper prior to optimisation; Inset 2: the boundary shape of the component; Inset 3: the final optimised shape.

Most importantly: “It is not about taking existing hierarchy’s and translating them. The evolution to 4IR requires a holistic and fundamental systems engineering rethink via upwardly scalable platforms that can harness the likes of Augmented Reality (AR), AI and Industrial Internet of Things (IIoT) capabili- ties,” he surmises. 4IR success requires flexibility What does this mean for organisations em- barking on a Digital Transformation journey? “The biggest reality is that in most cases companies don’t knowwhat they don’t know. Thismeans long-termplanning is out. Instead, companies need to provide for complete and multi-dimensional flexibility going forward, both in terms of the evolution of how data is deployed; as well as how collaboration is accommodated. “It is in this context that solutions such as PTC’s ThingWorx and Creo are vital to allow organisations this potential inanunchartered world going forward,” Teifel concludes. q

path canbeuse- fully collected now without limiting future expansion. “This rollout

is the biggest challenge. Do we explore areas aligned to our traditional business and imple- ment something thatwe knowwill worknow? Do we undertake a radical transformation of the organisation into a 4IR company? Do we create a brand new Greenfield organisation or dowe buy in solutions?Or all of the above? The decision has massive implications!” he warns. Teifel suggests focusing on the data evolution goal: to inform our specific busi- ness and industrial environment: “It’s about collecting, bundling and leveraging data relevant to one’s business environment,” he suggests. “Connecting sensible things from all parts of the spectrum, while maintaining complete flexibility to respond to the chang- ing environment.”

front-end company that dealswith customers directlywill focus on consumer and customer data,whileaback-endcompanysuchasamine will focus on its processes and performance. Organisations are not the same and the em- phasis needs to be different depending on the nature of the business. In some cases, a company’s reputation may be more valuable than its assets, for example, so this risk must be carefully managed,” he advises. In terms of technology for manufacturing companies, he says that autonomous robots, 3Dprintingandallthewaytoprocessautoma- tion using Internet-connected devices are all available. “Not all will be useful to everyone and incorrect adoption may kill a business. In terms of data, the rollout should be done so that whatever is needed on the evolutionary

January 2019 • MechChem Africa ¦ 19