– Kim Campbell, Chief Executive, EMA

The trek, which took place in June, was organised in conjunction with Callaghan Innovation, who partly subsidised attendance, and the Manufacturers Network (TMN) of Christchurch.

The anchor for the trek was the attendance at the Internet of Manufacturing Midwest Conference in downtown Chicago.

Also trade visits were arranged to Haas (America’s largest machine tool manufacturer in California) and Trumpf Midwest based in Chicago.

We also had a private seminar in Chicago and visited two business hubs, UI Laboratories and M Hub. The participants were a mixture of tech savvy New Zealand companies and one or two from Australia.

Connected Manufacturing (IoM)

1. AI and robotics have been around for a very long time and, contrary to perceived wisdom, have been slow to be implemented and are not moving as fast as the protagonists of technology would like you to think.

2. The potential is large. Proprietary enabling technology available through Alexa, Siri, Watson, Microsoft Azure and others has brought the entry points to very affordable levels for certain kinds of data interrogation. Retro fitting older equipment and connectivity are now possibilities for all manufacturers.

3. Outstanding and unresolved issues affecting both very large companies (Caterpillar, GE, Boeing) are cyber security, legacy systems and industry standards; all problems for which no one has a ready answer and are definitely slowing implementation. Related but solvable problem is in systems latency.

4. Data is seen as the “new oil” and the stats are staggering. By 2020 there will be 50 billion connected devices in the supply chain eco system. By 2020 predictive maintenance will be the killer app across all equipment (Smart Grid, Smart Cities, Smart Networks).

The supply chain optimisation through AI creates difficulties with vendor locking and creates tensions between proprietary and open systems.

5. 1% of all industry data is getting used (McKinsey) allowing for a 700% increase in productivity gains or $100 trillion in realised value by 2025. Data is growing at a combined rate of 40% per annum.

6. If these vectors are applied to maintenance costs alone, savings of 50% in outages and increases of 15% in output are possible.

7. The technology for mass customisation is already in place, but questions remain about how much of that actually will be driven by consumer demand.

8. Only 15% of companies have a discernible digital strategy leading to the prediction that 75% of S&P listed companies from 2012 will have disappeared by 2027.

The implications therefore are for companies to get on the programme, leading to the question, “What would a company who was to put you out of business look like”, and become that company.

9. The trend for manufacturing companies to move from “rules” based automation to the use of AI and machine learning to reduce variability in manufacturing so that each part is better than the last. This type of computing works best with low latency and edge computing rather than the cloud, where security and latency become an immediate problem.

10. The question of the societal shift and what happens when robots do all our work curiously remains unanswered, but here are a number of predications:

a) The first jobs to go are the ones described (dirty, dangerous and dumb) and the bumper sticker used for this is “shift happens”. The implications being we move quickly to eliminate the unpleasant jobs which are also dangerous, understandably for mining, cleaning, lifting.

b) Despite the predictions of mass unemployment, employment has actually increased and in most developed economies, there are skill shortages.

c) All this will take longer and cost more than anyone realises.

11. For manufacturing the issue is to attract young talented people who do not see the opportunities and have not been exposed to the modern factory. IoM uses a high level of technology and a low level of active labour engagement and is right in the millennial sweet spot.

12. The implication for New Zealand manufacturing is – rather than focus on big hit transformational changes, we should take early baby steps with simple sensors, post-automation and data integration. These use well understood proprietary technology. There are “killer apps” out there, but you don’t need these to participate in the digital revolution.

Temptation is to look for lots of “shiny things”, but really there is a need to find solutions to problems that are overwhelming the operation and seek to use IOT and to solving customers’ problems and not your own. This seems to me a real insight.

13. As a postscript, I visited Haas Automation. This is a non-union shop spread over 100 acres, with some 500 employees, with $1 billion in sales. Out of California they export 75% of what they produce which are components and robotic machine tools. Their underlying staff policy is employment at will, which effectively means they can hire and fire as they please. They have a waiting list for people to join the company.

They do not have a probationary period because everybody is permanently on probation. One wonders why anyone would work there until you find out that everybody gets an extra week’s wages every month that the production target is met and everybody shares in a profit at the discretion of the management.

The staff see no need for a union. The message clearly is if we are to have a flexible work environment within the technological future, then firms will need to be more generous in sharing the spoils of success.