For AI, good quality data is a necessity: ABB’s Guido Jouret
Mumbai: Guido Jouret, chief digital officer (CDO) of ABB Ltd, is responsible for the development and deployment of digital solutions inside the company, and for customers globally. A PhD in computer science, Jouret joined ABB in October 2016. He was earlier chief technical officer (CTO) at Nokia Technologies, prior to which he led the software products business at Envision Energy. Before joining Envision in 2014, Jouret was CTO and general manager of Cisco’s Emerging Technologies Group, and helped in setting up the firm’s TelePresence and Internet of Things (IoT) groups.
In a recent interview, he explained the role of a CDO in today’s world, the challenges therein and the opportunities that newer technologies have to offer companies. Edited excerpts:
Having worked both as a CTO and CIO and now CDO, tell us what exactly does a CDO do, and how is the role different from that of a CTO or CIO?
In some ways, the role of a CDO is to accelerate the company’s ability to adopt digital technologies, and develop new products and services. The reason it is different from that of a CTO and CIO is because a CTO tends to be more on the research side looking three-five years out, looking at enabling technologies whether they are electricals, material science or IT science, but maybe not so much focused on the here and now.
What changes have you introduced in these 18-odd months that you have been with ABB as CDO?
We have communicated our digital strategy to all our customers and partners with ABB Ability—now we have a common branding and messaging, and have published 500,000 articles on ABB digital in some shape or form. The second part of that strategy is that all our 21 business units now have a digital leader. This is the way we institutionalize our digital strategy.
We have also introduced a programme we call lighthouse projects to fund innovation. Business units submit project ideas that involve digital technology and if they meet certain criteria and goals, then I co-fund half of the budget required. We have had over 40 Lighthouse projects that have been funded, and 25 of those have already delivered a solution to the customer. Once we deliver to one customer, we can scale it and replicate it.
The other part that is very gratifying is our technology partnerships with companies like Microsoft, IBM and Hewlett Packard Enterprise.
Spell out some of the challenges you have faced while implementing this digital strategy.
Much will also depend on the business unit one is dealing with. What role, for example, will digital technologies play in a business unit that makes mostly electromechanical gear? Is it by putting sensors on these things (gears)? Is it by getting into services that are digitally enabled? Or is it by selling their products online? One of the ways we address this issue is with the Lighthouse projects. I use that funding to incentivize businesses to work together and collaborate even though that sometimes might slow down things initially.
Moreover, some business units need help with their business strategy. Other business units know what the business strategy should be—they just want access to platform technologies so that we can build these solutions more quickly, i.e. we do not want to build everything from scratch. For example: Do we have a cloud service? Do we have technology on-premise (in the company’s premise) in terms of gateway that they can use to build their solutions on?
What’s the experience with companies in India?
It’s not just the IT organization, or some CDO that is leading the charge (in India). I was pleasantly surprised to hear several CEOs describe their personal vision for digital technologies. Their vision statements are grounded in reality. The second trend is that you now see a mix of veterans and millennials in these organizations. That’s a very good sign because it is the pattern I am trying to promote within ABB. It reassures me that the approach we have taken is possibly a good one.
The benefit to be gained by using automation and digital technology in India is tremendous. For example, for per unit of manufacturing output India still consumes up to 30% more energy to do so. That basically says that if these companies became more efficient, the margin of improvement addressable is potentially as much as 30%. So you have quantum improvements that are possible.
When implementing digital technologies like automation, artificial intelligence (AI) and IoT, you advise clients not to get carried away by too much innovation. Please explain.
The necessary ingredient for taking advantage of AI is of course good quality data because AI today means mostly machine learning, and machine learning needs data from which to learn. So getting good connectivity, clean data sets and reliable data is a challenge for many companies today. Moreover, we have a lot of islands of automation. Hence, manual interventions not only slow down the automation process but are somewhat unreliable. In some cases, they are also not necessarily good for the finished product. For instance, in the case of food and beverage companies, one of the chief sources of contamination is the people working on the food. If you can automate some of that, then you can bypass this hurdle.
With increased digitalization, one of the main concerns is that of lack of standards and security. How should we address these concerns?
ABB has always been a proponent of standardization. We believe that adopting standards create more opportunities, even though it does mean that people can mix and match—they can choose some of your products and some of your competitors’. The adoption of Wi-Fi, for example, led to a dramatic expansion of the market. Ethernet is another example. However, we also know that the adoption of these standards in factories will take time. The reason is that if something is working properly, the replacement cycle could take decades.
That said, in the consumer space, Amazon has shown that if you interconnect devices on the cloud, you can achieve the necessary integration much more quickly. Because my Amazon Echo speaker is connected over the Ethernet and my Philips light bulb speaks Zibgy—inherently two different connectivity technologies—I can turn off my light bulb using Alexa.
The same cloud-to-cloud integration could increasingly be the way industrial solutions get connected. I can take data from the shop floor, send it to the cloud and link it to something else, which then in turn links it to something that I’ve got on my shop floor. Hence, I don’t have to wait for all these devices in the factory to be compatible.
On the security side, the long cycle time of products makes it more difficult to secure these products. Even if you have an updated product, there is no guarantee that the customer will update the firmware or make the change necessary. However, we do see a lot of customers begin to get interested in this area—particularly electrical utilities, automotive companies and others who believe that without this additional focus on cybersecurity, they are at risk.
AI is also improving the intelligence of robots, especially on shop floors…
That’s actually one of the greatest opportunities we see. So if you think about a robot today, first of all real robots don’t move. Most are nailed to a place and stuck in a certain position. Very few, if at all, have any sensors, which is why most of them are put in cages to protect the workers from the robot. On the other hand, a collaborative robot has plenty of sensors and is inherently safe. But teaching the robot a task is still a challenge. For example, we have to explicitly program a robot to know the difference between picking up an egg and picking up a bolt, failing which it could crush an egg if it tries to pick it up like an iron bolt.
If we have more intelligent robots, where they learn about the real world and properties of materials and objects, it would allow us to radically accelerate the time it takes before a robot can be productive. If these robots were endowed with more sensors and AI technologies, then we would see vastly more use of robots across sectors.
What are your thoughts on the impact of automation and AI on the workforce. On the one hand, there is a lot of scaremongering over machines dominating us and potential job losses. On the other hand, this trend can create new jobs.
AI has tremendous potential because of things like image recognition—look at the quality of cameras on smartphones, HD cameras, 2K cameras, 4K cameras where the sensing capability will continue to increase. That’s the equivalent of providing AI with better data even as the computational intensity increases and cost decreases.
The scary aspects of AI are felt mostly in the consumer space. In the industrial sector, it’s a space where we talk about improving safety, improving quality, improving the quality of the work—so in that sense, AI does not have the same sort of negative connotation that it might have in the consumer space.
We believe that more automation creates better jobs, and ultimately even more jobs. Look at the three countries with the highest level of the adoption of robotics—Germany, Japan and South Korea. They also have some of the lowest rates of unemployment.
These countries also, to some extent, have an education system that helps. They have traditional higher education of four-year degrees but have not neglected the fact there could be a need for tradesmen and craftsmen. This results in highly trained technical workers who can operate some of these increasingly complex (AI) machines. Just because you have a four-year degree, you will not necessarily know how to operate a 3D printer or a robot—these are specialized skills. We do believe automation will address the jobs that are dirty, dangerous or difficult.
In doing so, in the short term, it will displace some of those workers but in the medium term, automation and AI will lead to new opportunities as we are now creating new jobs. For example, who would have imagined that there would be people in 2018 whose job is to do SEO (search engine optimization)? So in many ways, it’s very hard for us to predict what the new jobs would be. The challenge we have is that we cannot yet see future markets today.
Do you see India equipped to handle such a situation, especially given it is a labour-intensive country?
It is indeed a labour-intensive country but there is also the incredible foundation of the IITs, which produce a lot of these top-level graduates in the ICT industry. One challenge is that the Indian economy has traditionally been more of a service economy than a manufacturing base. With initiatives like “Made in India” and the help of technologies like 3D printing, more manufacturing can take place at a small scale.
- Govt serious in bringing fugitive economic offenders to task: Rajnath Singh
- Sushma Swaraj arrives in China for talks with Wang Yi, SCO meet
- Make the best of technology to deal with administrative delays: Modi tells bureaucrats
- Amit Shah says ordinance shows Modi govt’s commitment to women’s safety
- Sanskrit most suitable for machine learning, AI: Ram Nath Kovind