Recent evolutions in data generation as well as machine learning - also called Artificial Intelligence for Manufacturing - will transform the manufacturing world radically, enabling Factories of the Future to move to ever more personalized and fully connected manufacturing.
Together with E.D.&A., Daikin, 24/7 TailorSteel, Van Hoecke NV and Newtec , Agoria and Sirris spent a full week in Silicon Valley to get a feel on the region's advancements in digitization & manufacturing. Following themes clearly stand out in today's Silicon Valley manufacturing DNA :
- Digital transformation : the process,
- AI (Machine Learning ) and
- Cyber security.
Today we'd like to share with you the facts and outlooks on the Digital transformation process theme of some local thought leaders.
Everything gets connected
Digital transformation in Silicon Valley's manufacturing community is mainly focused towards enhancing the customer's experience. Silicon Valley industry professionals are implementing the Netflix & Amazon.com principles, ie offering B2B customers access to
- 3rd parties,
- on-line Registration,
- payment and
- other application specific API's
through e.g. only one, unique identification action by the user of these services.
A second big trend is the fact that more and more virtualized systems & solutions are being offered by the ICT industry, enabling manufacturing companies to advance rapidly in (a combination of)
- remote monitoring,
- remote engineering and/or
- remote production.
First things first
The continuous ambition of manufacturing operations to decrease downtime and to improve OEE (Overall Equipment Effectiveness) spurs the willingness to engage in projects focused upon 'making manufacturing data become useful'.
Following important advice was given by several of our hosts during our trip : "Industry generates more data than any other business or market, but Industry also has one big extra challenge : the high VARIETY of data. This makes the build-up of intelligence as well as finding solutions much more difficult. Manufacturing companies therefore shouldn't engage in more advanced steps like predictive analysis, smart automation, energy usage optimization, etc when the basics are not yet taken care of. Problems are only getting worse if this data variety is not solved upfront."
For exactly the same reason, programs like Lean Manufacturing and 6 Sigma will continue to be extremely useful for getting these basics OK during the coming years. The company SightMachine presented an assessment tool for digital & technical readiness, incl. questions around the organizational set-up (eg KPI alignment between top and floor).
We heard quite some criticism on the launch of the Tesla Model 3 because of Musk's credo "let's go right to the cool stuff". This principle might be right for producing a few 100's of cars a month, but certainly not for 1000's !! - Ivo Bolsens, CTO of Xilinx, even stated that within 5 years from now, Elon Musk will not be making cars anymore…
Having put your foundation right, new platforms for multiple applications can be created, where systems continuously talk to each other. As a result, Silicon Valley's top manufacturing companies recently are engaging in broader, more business model focused transitions like the integration of smart products in production, creating microfactories, enabling capacity pricing, implementing manufacturing as a service, etc.
"It's not how info looks like on the production floor, it's about how it gets connected, how the systems talk to each other!" - Syed Hoda, Sight Machine (Manufacturing Analytics)
Digital transformation strategies of some large multinational ODM's (Original Design Manufacturers)
At the Global Manufacturing Innovation Centres of both Flex (formerly Flextronics) and Jabil (having also a production location in Hasselt), highly qualified manufacturing engineers people are mainly focusing upon
- Adding more sensors on the shopfloor
- Implementing predictive maintenance
- Introducing Augmented Reality (AR)
For their plants worldwide, Flex strategically chose following 6 future AR-applications :
- Business development & design
- Work Instructions & Training
- Remote assistance
- People & resources
- Services & repair
They started integrating these principles at their brand new Chicago manufacturing plant for producing game controllers. Please have a look at how this facility has been set-up :
"In order to move even more automated production (back) to high wage countries, the hourly cost is not the real issue, rather the established supply chain in low cost regions" - Ahmed Tantami, Sr Director, Technology / Innovation engagement Officer, Flex
3D printing : towards mature production technology
In a session @ GE Ventures, the Desktop Metal start-up company was introduced to us. Founded by an experienced team that includes MIT professors, serial entrepreneurs, and veterans of CAD software and robotics companies, Desktop Metal seeks to build office-friendly 3D printers that can inexpensively produce parts from a range of high-performance metal alloys. "Metal 3D printing has been out of the reach of most companies because it's very expensive and slow," says founder Ric Fulop. "We're developing a system that's fast and more accessible."
Main features of this GE-sponsored new metal 3D-printing technology include
- 100x faster than laser printing
- 20x less costly
- Bi-directional single pass printing ("whenever there is movement, there is printing")
Instead of working through programs like autoCAD, this new technology uses the principles of Generative design, supported by Artificial Intelligence. The user only has to put in the constraints of the product into the system and the computer will generate the product.
In the same session Mr. Paul Dilaura from Carbon3D presented a new, revolutionary type of plastic 3D-printing. Carbon3D was founded in 2013, is headquarted in California and completed its series D capital round. The Carbon company has more than 250 patents and patent applications. BMW, Adidas, J&J, GE and Nikon are amongst the investors.
Despite industry advances, traditional approaches to additive manufacturing force trade-offs between surface finish and mechanical properties. In contrast, DLS (Digital Light Synthesis) technology, enabled by Carbon's proprietary CLIP process, is a breakthrough technology that uses digital light projection, oxygen permeable optics, and programmable liquid resins to produce parts with excellent mechanical properties, resolution, and surface finish. A video illustration of this process can be found here.
Finally we want to share some of the conclusions of a debate organized by the FIT (Flanders Investment & Trade) office in San José.
The panelists were :
- Philippe Jansen Technical Solution mng, TSMC
- Paul Peeters Factory of the Future, Agoria
- Ken Caluwaerts Research Scientist, Google Brain Robotics
- Willem De Lange Technical Director, Synapticon
- Eli Reekmans Lead Engineer, Oqton Inc
Question : Within 15 years, out of every 100 jobs the 80 more mechanical ones will be replaced by machines. Which skills will be required for the 'jobs of the future'? In other words, what human capacity can never fully be developed by machines?
Answer : Repeat work will be mostly replaced. However, there is a clear need for new (engineering) skills to perform radical digital innovations, introduce digitally enabled process innovation as well as translating customer needs towards a solution.
Question: How will top manufacturing companies evolve the coming 5 – 10 years?
- Further merging and linking design and manufacturing
- Less departmental silos in manufacturing companies
- Improved flexibility, more collaborative co-creation
- Strongly increased competence to react on unexpected happenings
Our next blog will talk about Artificial Intelligence for Manufacturing, also called "Machine Learning". We'll keep you posted!
Did you miss our first blog on The Silicon Valley culture? No problem, you can still read it here.
Did you miss our second blog on IoT products for scalable business? No problem, you can still read it here