4. Shift towards an integrated digital and human workforce:
4.3 Reskilling for digital industries
The emerging job market will demand new and different skill sets. Digital-age skills such as data and analytics will become the “new math” in the Industrial Internet. To support an integrated digital and human workforce, society, educational institutions and business will also need to work together to instil a new mind-set on how to collaborate with intelligent machines, which, in some cases, may involve the need to teach and guide the machines as if they were apprentices. It is also important to recognize that machines, however intelligent they might be, are just tools. When in question, human experts must be ready to apply their critical judgment to overrule recommendations from automated systems.
Driven by constantly changing digital technologies, requirements and markets for Industrial Internet skills will be much more volatile. In response, learning and skill acquisition will need to be equally dynamic for individuals who seek employment. For instance, employees will need to perform more specialized tasks earlier in their professions, which will require them to regularly update skills through informal or independent learning, such as participating in massive open online courses (MOOCs). Instead of one-off degrees and technical courses, educational institutions will need to develop platforms for continuous learning, collaborating with students, businesses and governments to produce contents relevant to valued skills.
Job-related training and skills certification will become integrated into business processes and continuous, as there is more emphasis on delivering consistent outcomes and ongoing training across the extended enterprise. Such training will also reduce the length of onboarding time for new employees. Accenture research reveals that 79% of organizations already use just-in-time and social learning to build skills quickly.38 For example, a newly hired retail sales associate could be given a wearable intelligent assistant on the first day of the job. When a customer asks a question about a product, the tool would use automated speech recognition to detect verbal cues, and deliver relevant product information. This just-in-time delivery of information could enable the associate to learn as he is helping the customer.
The same quantification capabilities that power the outcome economy will also be at work in skills development. Employers will use cognitive training to develop detailed models over time on how workers think and act in specific job situations. Using this data, companies can tailor training programmes to individuals to make them more effective and efficient. Such accelerated learning techniques also offer great potential to align training and skills with content and context, as successfully demonstrated in a military setting. For example, as part of the Accelerated Learning Program, the US Defense Advanced Research Projects Agency (DARPA) uses neuroscience principles to improve sensorimotor and cognitive functions.39
The workforce impact of digital technologies will be gradual and profound, as the Industrial Internet transforms industries and business practices. Because system-wide changes take time and planning, business and government leaders and planners will need to act now in preparing for the digital talent market. Some of the initial steps might include examining existing approaches, experimenting with new digital workforce models, and developing a comprehensive strategy on how to reform the education and training system to be more responsive to the demands of the future workforce.
The future of robots
Robots are microcosms of the Industrial Internet. They feature three core capabilities: sensing, thinking and acting. Most industrial robots used in manufacturing today are no more than advanced control arms with limited sensing and reasoning capabilities. Just like the machines around them, these robots are preconfigured to carry out repetitive, structured tasks.
As sensors, hardware and software continue to improve, robots will become more intelligent and autonomous in their capabilities while still working under human direction. For instance, robots will eventually be able to understand the physical world around them, in much the same way as humans do. As a result, robots will appear freely in open environments, such as offices, homes and shopping malls, doing tasks that only humans once did. The use of service robots is expected to grow faster than the use of industrial robots in the near future (e.g. security “guards” from Knightscope).40
One distinct capability in the next generation of industrial robots, such as Baxter or Universal Robots, is their ability to work safely alongside humans. New sensors and software enable these machines to detect and avoid collisions with people, and such robots are now also reprogrammable so that they can quickly “learn” from human workers how to perform new tasks. These features, together with lower costs, mean that robots will be deployed more widely. As human co-workers, collaborative robots are likely to reshape manufacturing processes and workforces. In bringing automation to new applications, robots could also help manufacturers in high-cost countries regain a competitive edge, which might also mean fewer jobs for the lower-skilled workers but more higher-skilled jobs instead.
At the technology level, robotics represents one of the most exciting areas of innovation among corporate R&D labs, start-ups and university research centres. Here are just a few examples:
- Qualcomm is designing a new brain-inspired chip called Neural Processing Units (NPUs), which will be both highly scalable and power efficient. The new chip promises to redefine the cost/performance ratio for robots, just as mobile chips have done for smartphones.41
- Google continues to advance machine learning by acquiring a series of robotics and artificial intelligence start-ups.42
- At Cornell University, researchers are building a large-scale, cloud-based knowledge repository called Robobrain, which can be used over the Internet to teach robots like Baxter how to comprehend (sense) their environments and quickly take on new tasks.43
Similar to the advancement of mobile technology 15 to 20 years ago, the robotics revolution is just beginning. Over the next 20 years, it will likely lead to profound impacts on businesses, the economy and society.