Exclusive: Leading food manufacturers reveal how they harnessed the power of Artificial Intelligence

Recently, I had the pleasure of hosting a roundtable discussion with senior executives from six leading global food manufacturers, on the topic of “Artificial Intelligence and Food Manufacturing - Opportunities, Challenges & Strategies”.

The panel took place at the Industry 4.0 Leaders Summit, and included senior executives from Nestle, Pepsico, Mondelez, Danone, Bimbo and Aryzta - so it was a really exciting opportunity to hear firsthand how leading food manufacturers are approaching such a hot topic.

The panelists all shared their own unique experiences of adopting and leveraging Industrial Artificial Intelligence to gain a competitive advantage. As you can imagine, there were a lot of interesting insights and recommendations, particularly for other manufacturing executives who are thinking about embarking on a similar journey.

In this post, I want to list some of the key takeaways (You can watch the full recording here.)

Artificial Intelligence is just a means to a goal - focus on the goal!

Our first panelist - Eugenio Alvarez, Associate Engineering Director at Mondelez - put his finger on a common mistake when it comes to adopting new manufacturing technologies, and suggested a simple way to avoid it:

“Many people consider AI and digital solutions is like a journey to get to, (but) for me it's a journey to use - it’s a means and not a goal.

So we'll always say first connect with the business and link the technology where the business needs it most.

This was a theme that almost all of our panelists emphasized. The successful adoption of exciting new technologies like Industrial AI relies on one thing: meeting your strategic business goals.

It sounds obvious, and of course most executives know this intuitively (nobody is out there simply searching for “cool technology” to buy.) But quite often when these kinds of “innovation” and “digitization” projects do hit a snag, it’s because somewhere along the line the technology simply isn’t right for the business problem they’re trying to solve.

Sometimes that’s because the technology doesn’t live up to the promise -- but at least as often it’s actually because the business case wasn’t clear enough from the start. Once the business case is crystal clear and focused, it’s a lot easier to zero-in on the right technology.

As Lisa Gofna Rubin, Open Innovation Manager at Nestle-Osem put it:

“It's not about testing a new technology - it's about testing a technology that we know upfront will make a difference for the business.”

The only way to know that “upfront” is to go into the process with a crystal-clear understanding of the business needs you want to address.

Here at Seebo, we actually developed a simple methodology for matching a specific manufacturing challenge to the right Machine Learning / Artificial Intelligence technology. We call it the Industrial AI Quadrant, and you can learn more about it here.

Why manufacturing giants hook up with startups

Here at Seebo we’ve been privileged to work with and learn from some of the world’s largest manufacturers. In the food industry that includes the likes Pepsico, who have an entire organization (Pepsico Labs) dedicated to sourcing and working with startups to accelerate innovation.

Of course the immediate question is why? What does a giant like Pepsico have to gain from startups like Seebo?

Anna Farberov, Head of Global Venture and Innovation at Pepsico Labs, offered an insight that many other large manufacturers can surely relate to:

“We really didn't have a choice, startups are driving the pace of innovation. To be innovative, to be really out there and to develop as a company, we need to be very good at working with startups.”

That recommendation did come with a word of caution, however, as Anna emphasised how -- just as with selecting a technology -- it’s important to carefully consider which startups to work with.

"There are many startups out there, many amazing technologies -- but many of them are very early-stage. So they won't necessarily have the traction and the experience. So how do you tap into the global ecosystem of startups and how do you make sure that you have access to the best of the best? You're not wasting time on startups that will not succeed eventually or will not be able to deliver eventually?"

Her answer? Again, it comes down to focus:

"...Focus on the right business needs, focus on the right business teams, stay close to the problem and partner with the people who will be executing the solution, because it's very easy to stay in your corporate ivory tower and that know how this is going to be implemented and this is how it's very easy to fail."

How to make an “excellent” production line even better

Technologies like Artificial Intelligence are often viewed as a solution to a painful business problem.

This is certainly the most common use case we see among our customers, across many industries. For example, a production line with particularly high losses, or a line with serious capacity issues. Of course, this is a perfect scenario in which Industrial AI solutions can help, by revealing the hidden causes of losses and process inefficiencies -- as many of our panelists noted.

But AI can also be applied to production lines and factories that are already performing above average -- or even exceptionally well -- in order to take them to the next level of process efficiency (what we at Seebo often refer to as “process mastership”).

A number of our customers use the Seebo solution to do precisely that. Once their existing resources have been successfully employed to squeeze out as many process inefficiencies as possible, they use Seebo to eliminate the most stubborn inefficiencies and losses that still remain.

This is the case with Danone, whose Senior Process Manager, Rene Waalewijn, described his journey with Seebo -- and offered three tips to other manufacturing executives on how to “think outside the box” and reach new heights of process efficiency.

“First of all reconsider your basic assumptions. By doing that you can take the process efficiency to the next level. Secondly, look at the market needs and requirements; ask yourself, is there a better, faster, cheaper way to deliver them while still providing, of course, a superior product to the markets? And then last but not least, work with a vendor that cares about those needs as much as you, don't look only to technology but also look for a solution to a problem.”

Artificial Intelligence is already driving manufacturing losses reduction

Of course, one of the most significant takeaways from this roundtable was the fact that all of these major manufacturers are already using Industrial Artificial Intelligence. Each of these household names are adopting and incorporating AI in their day-to-day operations, to address some of their toughest business challenges.

And they’re succeeding. They’re reducing waste, improving quality levels and increasing yield and throughput - which is why these companies, along with many others, are investing increasingly more on Industrial AI technologies

For more fascinating insights, watch the full roundtable featuring thought-leaders from Nestle, Mondelez, Bimbo, Pepsico Labs, Danone and Aryzta:


The Internet of Things - smart product adoption is on the rise in 2016

The Internet of Things: 5 signs smart product adoption is growing

2015 was the hype year for the Internet of Things (IoT), or what research giant Gartner refers to as the ‘peak of inflated expectations’. Despite strong predictions by Gartner and others IoT studies, people wondered if the Internet of Things would lead to genuine smart product adoption in the mass market. 

A year later, the results for 2016 are in: While adoption is progressing slowly in some industries, smart products have led to increased revenue for numerous verticals and raised expectations for ROI from the internet of things. Market reports go a step further and now declare that IoT is "going mainstream". 

If 2015 was the year of Internet of Things hype, here are five signs that smart product adoption by consumers, manufacturers and ecosystem players is on the rise.

1. Smart product adoption is generating ROI

The Internet of Things brought in $750 billion worldwide in 2015, according to a Machina research report. A growing number of companies already reported increased revenues thanks to smart products integrated with information systems and devices. Intel, one of the largest stakeholders in the Internet of Things, saw its IoT revenue grow by 2% to $572 million this past year. Glancing at a few verticals, sales of smart home audio devices and ‘hearables’ brought the revenue for the smart audio market to $1 billion this year. Similarly, in the industrial world, smart electricity meters generated over $4 billion in revenue in 2015. The growing adoption of smart products has prominently impacted revenue for everyone in the IoT ecosystem, and companies are vying for greater roles in the smart product market in 2017.

2. Big companies are fighting to strengthen their footholds in IoT

Companies like Cisco and General Electric are changing strategies and fighting for dominance in the Internet of Things. Cisco intends to lay off nearly 7% of its workforce – 5,500 jobs – in a restructuring that includes new investments in IoT, next-generation data center, cloud, and other areas. The company saw an increase in revenue in 2015, strengthening their claim that the move is a proactive one. It is even better understood in the context that Cisco and General Electric are racing each other to “dominate the widely coveted Internet of Things landscape.”   The two giants are now in a dead heat to re-brand themselves as premier industrial IoT providers.

The dramatic changes occurring in Cisco are a clear sign that communications and computing giants consider IoT to be the best investment they can make for their company. And that's a good sign for smart product adoption for other companies as well. 

3. Smart product adoption by IoT players: designers and developers are getting on board the Internet of Things

90% of designers surveyed by Morgan Stanley are “adding connectivity to IoT.” The company’s AlphaWise survey in May pointed out that software developers and product designers who want their skills to stay relevant, look at the way the industry progresses and sniff out opportunities. This data correlates a survey from Upwork, which saw a dramatic increase in the number of job postings for learning IoT-related development skills. VisionMobile projects there will be 4.5 million developers involved in IoT by 2020, "a 57% compound annual growth rate and a massive market opportunity". Smart product design and the accompanying software and firmware development skillsets are high in demand, and professionals are broadening their skillsets to adopt to the new technology and advance with it. 

4. As competition increases, variety flourishes, disruption status disappears, and users buy more

The smart product market is widening. One smart offering will disrupt an industry, but ten will create new standards, raise competition, lower prices and introduce a diversity of new features. Smart sprinkler controllers now set the standard for the industry, thanks to big names like Rachio and Skydrop. There are also whole lists of smart irrigation controllers for commercial and industrial use. The same applies for smart thermostats and air conditioners, which, according to Navigant research are some of the most popular smart products being purchased by consumers. Nest, Sensibo and Honeywell show that everyone from established brands to spunky startups are increasing their revenue from smart home products.

A perfect example is the smart vacuum cleaner industry. Just as smart products in 2016 see more competition, forcing manufacturers to get creative with smart features, competition drove the robotic vacuum company iRobot to reinvent Roomba, the original robotic vacuum. through IoT. Today, Roomba operates remotely via an app on smartphones. The same product that was once a disruption is now removing the disruptive status of smart products.

Smart product adoption by industry
Projected smart product adoption for self-driving vacuum cleaners and other products in the next few years. Image Source: accenture.com, 'The Future of Consumer Adoption'

5. Short range connectivity solutions are a growing, viable option

According to the Machina Research report, “71% of IoT devices use short-range [connectivity solutions] such as Wi-Fi and Bluetooth.” Consumer electronics, building and automation are the top verticals taking advantage of these solutions. However, component prices have shot down in the last few years and the supply and competition between different solutions have increased, all making connectivity a more viable option for consumer goods and industrial manufacturers.

Numerous reports are already coming out with positive IoT revenue forecasting for the next five to ten years – everyone from Gartner and IDC, to Berg Insight. But a look at data from 2016 alone is promising: More smart products enter the mainstream market and drive business value for everyone from startups to large corporations; more designers and developers learn the skills necessary for participating in the IoT ecosystem; and connectivity solutions are democratized to enable more product companies access to wireless connectivity. 


IoT white paper - smart products transform business growth


Unlock Business Value with the Internet of Things

The Internet of Things: A New Way to Unlock Business Value

With the advent of smart technology, there is an astounding breadth of new capabilities that IoT technology enables for consumer products. A new generation of smart products allow for integration of services, either embedded into the physical product or offered via a connected application on a smart device. With products and the consumers who use them so greatly changed, business models must take into account that there are new needs to answer and new goals to achieve – but these goals and needs must be met through new models.

In order to ensure business value from your smart product, you should first explore how the IoT business model creates, delivers and captures value in an entirely original way.

The Internet of Things is all about turning products into services

The product is larger than its physical dimensions and includes digital services that create value during the in-market phase. Take a smart bike, for example. Bike theft is a serious issue for anyone who cycles frequently. A smart bike with GPS tracking connected to an application is incredibly valuable for owners, who can instantly identify the bike’s location and retrieve it. But once the smart bike is connected to an application, it accrues value exponentially to cover much more than theft prevention. The same app serves as a social tool that enables multiple users to connect with one another, suggest routes, swap advice, and plan trips together. Another possible service would connect users to the nearest bike repair shop, directly on the app. Deciding which features to give your product and thus extend its value depend largely on your ability to think outside the box.

Revitalize your product and keep it fresh for users. Extending the product lifecycle is one way to extend your smart product's potential value. Once a product goes on the market, it has a figurative ‘best-by’ date, at which point it becomes obsolete and its value runs out. With smart products, the ability to update the product software remotely, and in a centralized manner, revitalizes the product and keeps it relevant for a significantly longer time.

iot business value model
Value creation and capture, traditional product vs. Internet of Things product. Image credits: hbr.org

To get strong business value, rethink value creation: Find opportunities for services and create new habits

If up until now manufacturers responded to customers’ needs retroactively, with the Internet of Things it’s about predicting new needs - or even inventing them. Today you can establish new user habits to create new sources of value.

Going back to the bike example, adding GPS tracking that connects with a smart device enables owners to find stolen bikes easily – or locate where they parked it. Owners can activate the camera on their smart bike, with a tap on their smartphone screen whenever it’s parked, to catch a thief - or turn it on before a ride, to act as evidence in case of a traffic accident. Safety in general is a major concern which smart products tap into; other examples include smart infant and child car seats that alert parents if children are left in a locked car, and smart doors and windows which automatically alert owners if someone attempts to break into their house. By providing new answers to old problems, smart products can create new standards of value in their industry.

The IoT Ecosystem: Identify who can best deliver each source of value or services

Within the Internet of Things, you don’t need to develop everything on your own. The trick to turning products into services is to create an IoT business model with an ‘ecosystem approach of doing business”; i.e. integration and partnering with outside stakeholders. In the realm of smart products, “value creation” occurs inside of “an ecosystem with multidirectional value and service streams”.

There are numerous “stakeholders”, or players, in the IoT ecosystem.  Product companies need to engage with these stakeholders to complete the layers that lead to value creation and capture: technology skills, hardware components, IP rights, resources, and other things that provide a competitive advantage.  App and hardware developers create business value through innovation, and “the business ecosystem” captures value during the in-market phase. The connected app, for example, enables you to suggest add-ons and other digital services that give substantial value, many of which will come from partners in the IoT ecosystem. Once you know which business values you want, ask who will add them – and which stakeholders can best deliver them.

It’s equally important to understand exactly what the incentive is for each stakeholder, in order to encourage “sustainability of value creation within a stakeholder network”. It’s a two-way approach: The ecosystem must deliver value to customers, and in return, each player in the ecosystem benefits from their involvement. Provide the right incentive, and you guarantee the best possible service and delivery of a source of value from their end.

iot business value depends on a strong ecosystem of shareholders
The Internet of Things gives the best business value when there is a strong ecoystem in place to deliver services.

4. Validate services and other new sources of value through user data

With IoT, the black box that prevents product companies from understanding product-user interaction and user experience disappears. Instead, real-time analytics shed light on the customer’s interaction with the product.

User data provides insight that enables product companies to prolong the product lifecycle by optimizing popular smart features. With this astonishingly valuable insight, manufacturers discover exactly what value the product holds for their users: which features they use, where, and when.

Returning to the smart bike, if real-time analytics show that users tend to bike together, the product can feature more social functions on the app, or the firmware can be reconfigured so the bikes ‘recognize’ one another. The possibilities are endless; it’s a matter of choosing which data to analyze and how to handle the subsequent insight.

Business Value in the Internet of Things can be summed up in one word: opportunity. With the new potential to anticipate needs and excite consumers, smart products empower product manufacturers and companies to anticipate or even invent needs to excite and surprise consumers, and make the product as necessary to everyday life as smartphones and computers. They also enable the leveraging of in-market insight to find new business value and enhance the product while in use. Most importantly, having a strong ecosystem of third players who deliver services and sources of value provide companies with the optimal smart product and the highest business value from IoT. While some of these value drivers may not come into play until the product is on the market, grasping the importance of these elements when planning a business model guarantees a much greater value from the smart product in the long run.




Build smart products and speed up time to market  - the IoT Platform with virtual automated tools