The long-term disrupter
Automation first became a mainstream phenomenon in the late 1970s when robots displaced manual functions in the assembly of cars and the manufacturing of packaged food and beverages. Why then, almost 50 years on, do we still consider automation a valid and disruptive investment theme?
The answer lies in the combination of technology and communication in automation hardware that continues to expand addressable markets as well as improve the quality and efficiency of sectors where it is already well entrenched. Even the most conservative forecasts suggest the market for automation will grow by 9% per annum and represent a $1.5 trillion revenue stream in five years’ time. However, just as society’s enthusiasm to adopt automation is likely to waver, we can expect the investment opportunities to shift between sectors, services and technologies.
From hardware to software
Take robotic arms for example. These machines continue to improve upon the efficiency of repetitive discrete functions, and remain a key part to improving the quality, safety and efficiency of early adopting industries like auto manufacturing1. However, technology now enables robots to expand into bespoke processes such as those that require a real time sorting of random shaped and textured products, spaced at irregular intervals. This has enabled companies like Ocado to commission an automated system that can differentiate, handle and package a mixed basket of groceries and household products. It has also given rise to ‘cobots’ which as the name suggests, are robots that are able to work alongside humans. Although the early movers have been industrial processes, the medical and consumer sectors are also realising the benefits of integrating co-bots in certain procedures as well as patient monitoring and care.
As automation evolves, so too does the value chain. The hardware components and their assemblers have been unsuccessfully battling the headwinds of overcapacity and retaining intellectual property. Basic robots are now largely commoditised. Fanuc, a Japanese robot manufacturer, famously has a production facility in which robots produce robots. However, the consequent price deflation has enabled lagged sectors such as textile and materials handling to adopt automation where the efficiency gains have been until now less readily quantifiable. As technological sophistication increases so too does the portion of value captured by the enabling software. The automation players have spent the last decade, and a considerable amount of capital, building out software capabilities that leverage the scope of their hardware and simultaneously ties in the customer.
The relevance and evolution of automation has also been driven by communication. ‘Dumb metal’ (the unflattering term bestowed on commoditised discrete process robots) has become intelligent; communicating through control functions or artificial intelligence learning (AI), automation functions can adjust to a change in the operating environment, predict their own hardware failures and even ‘learn’ to further improve the efficiency of their own function. Importantly, in combination with technology, it has enabled the benchmarking of different vendor machines and process facilities on a real time, like-for-like basis. This opens up the potential for a further incremental step change in efficiencies, quality and safety. For green-field sites, the concept of a ‘Smart Factory’ in which automated production is flexible, self-optimising and quality assured has become a reality.
Industry Structure
Technology and communications will sustain the evolution and relevance of automation for the next fifty years, which suggests it is almost inevitable that we will see new entrants into the space and some of the existing players lose their relevance. We have already witnessed the commoditisation of dumb metal, but the consequent price deflation is not necessarily a recipe for poor returns if you are the leading, lowest cost manufacturer. The last man standing in a high volume growth product can be a winning formula.
As already mentioned, in recent years the value has benefitted software but as the automation providers shift from closed to open platforms, the vendor pricing becomes more opaque and the customer’s loyalty arguably less robust. It is becoming more challenging to justify the acquisition of software capabilities now at valuations that are two to three times the multiple of the incumbent hardware manufacturer.
As automation touches more sectors of the economy and globe, its gearing to industrial capital spending cycles will decrease. Perhaps a broader customer mix will mitigate industrial cycles and lead to a superior valuation, or will investors prefer to shift between sub-sector specialist providers? Where we can be more certain of accruing value is where the industrial structure has and will continue to allow pricing power. This is proven in the optics sector where a handful of companies with strong innovation cultures (and spend) are providing increasingly sophisticated products to a structural growth market. Co-bots need to ‘see’ more clearly, everywhere; automated production, sorting and packaging need a greater degree of optical monitoring to ensure quality. These mission critical optics represent a small portion of automation products and projects; vendors are reluctant to switch to a new entrant with unproven product performance, weaker innovation and service support even if it costs 20% less. It could be argued a similar virtuous structure prevails in automated lasers.
Future risks and opportunities
Automation efficiency and optimisation are often euphemisms for labour displacement. Robots only partially service the requirements prompted by an ageing population profile and the unskilled labour force is disproportionally threatened. The World Economic Forum report2 estimated 5 million people will lose their jobs to robots in the next three years. Bill Gates in particular is a growing voice calling for a tax on robots that is deployed to re-skill the displaced and contribute to social welfare.
Labour displacement in automated defence applications also has an ethical angle, but one that centres on the implications of remote warfare. The integration of stewardship in our investment philosophy and process ensues we assess the ethical headwinds in a robust and quantifiable manner, not just to reflect current exposure but also to capture future potential risks.
The compounding growth in automation adoption underlines its continued relevance as a mega theme. The past 50 years has demonstrated significant volatility in the pace of adoption and the dynamic sweet spot of the value chain – and this is unlikely to change. We will remain cognisant of the technological evolution of automation to a broadening customer base that will likely throw up new entrants and challenge the established hierarchy. Automation in my world may calculate the algorithm of an automation and robotic ETF, but it is highly unlikely to identify the new entrant or incumbent winners.
1Discrete manufacturing is the production of distinct items such as automobiles, furniture, toys and smartphones.
2World Economic Forum. ‘The Future of Jobs – Employment, Skills and the Workforce Strategy for the Forth Industrial Revolution’ January 2016.