00:00:03 Internet of Things (IoT) and China’s smart cameras.
00:01:05 Cheap computing power enables IoT in everyday objects.
00:02:24 IoT’s impacts: enhancing device functionality and supply chain.
00:04:47 IoT’s role in global supply chain management.
00:07:02 Cheap IoT devices: power supply and sleep modes.
00:09:46 IoT data integration challenges in ERP systems.
00:11:28 Amazon overcoming IoT data-related challenges.
00:13:27 Future: Automated activities with IoT, e.g. Amazon Dash Button.
00:13:38 IoT adoption challenges: security and botnet creation risks.
00:16:02 Servers’ struggle to differentiate human/machine requests.
00:17:12 IoT devices: Attractive targets for cyber criminals.
00:17:49 IoT security solutions: Code auditing, hacker bounty programs.
00:20:55 Future of IoT: Autonomous vehicles, supply chain logistics.
00:25:17 Organizational/IT changes for increased IoT usage.

Summary

In an interview on Lokad TV, Joannes Vermorel, Lokad’s founder, speaks with Kieran Chandler about the growing prevalence and implications of Internet of Things (IoT) technology. Vermorel underscores the potential of IoT for revolutionizing supply chain management via real-time asset tracking and environmental monitoring. He also recognizes the affordability and energy efficiency of IoT devices, while warning of the significant data management and security challenges they can present. Vermorel points to Uber and Amazon as firms skillfully utilizing IoT, with Uber’s driver tracking and Amazon’s intelligent supply chain operations. Looking ahead, he anticipates a rise in IoT-driven supply chain traceability, despite implementation obstacles in IT and organizational structures.

Extended Summary

In the latest episode of Lokad TV, host Kieran Chandler and guest Joannes Vermorel, Lokad’s founder, explore the subject of the Internet of Things (IoT). Vermorel discusses the ever-growing presence of IoT devices in our interconnected world, a trend largely fueled by the falling cost of computing power. This affordability allows the integration of computers into almost any item without a substantial price increase.

Vermorel refers to cars, home temperature control devices, IP cameras, and smart home devices like Amazon Echo as examples of items that can gain IoT capabilities. He states these enhancements make objects more versatile with features such as self-diagnosis and network connectivity, paving the way for smarter object maintenance, preventive repairs, and simplified usage.

Much of their conversation concentrates on the possible impact of IoT on supply chain management. Vermorel suggests that IoT can play a vital role in tracking and monitoring goods from manufacturing to their final destination. This can help tackle key supply chain challenges, like coordinating a large number of moving assets over extensive geographical areas.

By integrating IoT into objects like pallets or parcels, Vermorel points out that real-time tracking becomes possible, allowing for better asset management. Moreover, IoT devices’ role in supply chain processes extends beyond merely tracking objects’ locations. IoT sensors can capture and transmit environmental data such as temperature and acceleration, ensuring goods’ security and monitoring their condition during transportation.

The affordability and energy efficiency of IoT devices is another key topic of their discussion. Vermorel illustrates how decreasing computing costs make it viable to add such functionalities to almost any item. The cost of embedding a fully functional computer with internet capabilities into a product could be as low as single-digit dollars, turning IoT devices into virtually consumable items within the supply chain.

Vermorel explains that IoT devices, leveraging advancements from mobile technology, can conserve energy by going into a sleep mode when not actively processing or transmitting data. These devices can operate for one to two years on their battery, eliminating the need for external power. They can be viewed as disposable, with their end-of-life disposal and recycling incorporated into their lifecycle.

They also discuss the challenge of integrating data produced by microprocessors into existing ERP systems. Vermorel acknowledges that IoT devices generate a massive amount of data. Although each device produces a small amount individually, the aggregate output from many devices is substantial, necessitating new IT strategies such as NoSQL databases and big data capabilities for effective data processing.

Vermorel cites Amazon as a company that has successfully tackled this issue. He commends Jeff Bezos’s long-term strategy and Amazon’s service-oriented architecture. Amazon’s acquisition of Kiva Systems, a company specializing in warehouse automation through robotics, and the launch of Amazon Dash buttons showcase Amazon’s innovative use of IoT.

Vermorel then identifies security as a significant challenge to IoT adoption, alongside data management. He asserts that IoT devices are more susceptible to hacking than conventional data center computers, which could result in these devices being hacked and incorporated into a botnet for cyberattacks. He uses IP cameras as an example, which are frequently hacked and used to create some of the largest botnets.

Next, Vermorel introduces the concept of botnets to the audience. Controlled by criminals, botnets can be employed for various malicious activities, most commonly denial-of-service attacks. These attacks use the botnet to overwhelm a website with traffic, blocking access for genuine users. Criminal groups often exploit this as an extortion method, offering to halt the attack for payment. IoT devices, due to their weak security and internet connectivity, are prime targets for such attacks.

The complexity of securing IoT devices is the next topic Vermorel addresses, comparing this challenge to security breaches experienced by established companies like Intel. With no one-size-fits-all solution, Vermore

l emphasizes the importance of regular source code audits and stress-testing devices to identify vulnerabilities. He suggests that ‘bounty programs’ from technology vendors could encourage ‘white hat’ hackers to discover and report potential breaches, thus improving overall security.

Vermorel then turns to the potential and challenges of IoT. He envisions a promising future, acknowledging the existing hurdles. He mentions that Uber, as an example, already utilizes IoT on a large scale, tracking every vehicle’s position in their network through drivers’ smartphones, providing their dispatch system with extensive control and efficiency.

Vermorel also emphasizes Uber’s transition toward a network of autonomous vehicles. He anticipates that, once autonomous vehicles become mainstream, Uber will essentially operate as a massive IoT network controlling moving assets. He adds that autonomous trucks are beginning to appear, with Uber launching the first fully autonomous truck line in the U.S.

The discussion concludes with a focus on supply chain traceability, where Vermorel sees IoT as a powerful tool. It can offer real-time updates and full traceability, crucial for industries dealing with high-value goods like consumer electronics, luxury items, or vaccines. Vermorel acknowledges Amazon as a pioneer in using IoT for supply chain operations, adding agility and flexibility to the process.

By leveraging real-time insights down to the pallet level, Vermorel suggests that companies can make intelligent, instantaneous decisions, like rerouting a delivery in progress due to a sudden need. While he predicts that these applications are on the horizon, he warns that implementing changes in IT and organizational structure to accommodate new operational methods could present significant challenges. However, he reassures that, if implemented correctly, a dystopian ‘Black Mirror’ scenario is unlikely.

Full Transcript

Kieran Chandler: In today’s episode, we’re going to discuss everyday devices which, when connected to the Internet, become smart. These go by the rather ambiguous title of the Internet of Things. Nowadays, we’re living in a truly interconnected world where the range of devices being embedded with software and sensors is seemingly growing by the day. This increase in connectivity undoubtedly has its benefits. For example, in Shenzhen, China, the installation of 40,000 smart cameras across the city saw the crime rate slashed by over 50%. However, these everyday devices can be hacked. At best, the hackers might be using the device to glean information for targeted advertising campaigns. However, at worst, these loopholes can be used for service-level attacks which have been used in the past to take down entire countries’ banking systems, such as what happened in Estonia in 2007. So, Joannes, with a buzzword like “Internet of Things”, it’s very vague sounding. Perhaps we can start things off with a bit of an example here?

Joannes Vermorel: Yes, absolutely. The Internet of Things is just a trend of computing power becoming cheaper over time. A couple of years ago, it became cheap enough so that you could have a computer in your pocket - that’s your smartphone. The trend is still ongoing, and now it has become so cheap that you can actually put a computer on pretty much anything. And for many things that are not super cheap already, it will barely change anything on the price of that thing. For example, we can put computers on cars - cars are pretty expensive, so they can contain literally dozens of microprocessors and smart processing power already. We can also put them on devices that control the temperature of your home, on IP cameras that only require an internet connection to upload their video stream to an online service of your choosing. More recently, we have seen a trend with smart devices like Amazon Echo, where you can just say a voice command out loud, such as “buy me this song”, and it will proceed directly with taking the order and processing it.

Kieran Chandler: If we look at it from a human perspective, how is this technology really changing how we’re interacting with these objects?

Joannes Vermorel: I think it can make objects more capable in many ways. They can have self-diagnosis capabilities so that if you want to do a repair, many objects can provide a readable diagnosis if something goes wrong. For instance, most printers will display the problem directly on the printer. Most printers can now be attached to a network, and then whoever has access to this local network can print on the printer. You don’t necessarily need to plug the printer into your computer. The same logic can be applied to enterprise coffee machines, which might require repairs. These machines are capable of self-diagnosing and triggering a maintenance operation on their own instead of waiting for the problem to happen. It’s not just objects that were already powered that can be augmented with IoT capability. From a supply chain perspective, we can add IoT capabilities to objects that were not typically self-powered, like pallets or parcels, just so that we have better tracking systems in place that can directly feed information to the headquarters about the status of this particular object.

Kieran Chandler: So, you’ve mentioned pallets and parcels and trucks, which is looking at things from a supply chain perspective. Where do you see the real potential for supply chains with the Internet of Things?

One of the biggest challenges of supply chains is dealing with a world that is vast, with plenty of assets that are moving and need to serve clients’ needs. It’s very hard to have tight, even real-time coordination on everything at a global scale. So, IOT offers a possibility for active tracking on every single pallet, truck, parcel, and we know exactly where they are and where they’ve been over their entire lifecycle.

Joannes Vermorel: Yes, it’s not just about having the location. When you add an IOT tracker, you can track temperature so you know if the temperature has been controlled for the entire duration of the transportation. You can track acceleration with an accelerometer so you know if your goods have suffered shocks that could damage them. You can also have smart security measures to monitor the physical integrity of a seal, to make sure that what is being transported hasn’t been tampered with. These capabilities are already being used today with simple, less technological approaches, but with IOT becoming super cheap, the question becomes, what can I do with this computer that I can attach on pretty much everything?

Kieran Chandler: That’s fascinating. But something I’m really sort of missing here is where’s the external power coming from for these microprocessors and microcomputers being attached to every single one of your pallets. Don’t they need to be charged up?

Joannes Vermorel: That’s a good question. There is a very clever trick that has been inherited from the progress made on mobile phones. Your IOT device doesn’t need to be powered all the time. It can go in sleep mode and then wake up, say once per minute, do some processing for a tenth of a second, send a pulse across the network, and then go back to sleep. When you do that, you can preserve the energy of your battery for a much longer period of time. Modern IOT devices, if you don’t demand too much from them, can last for one or two years.

Kieran Chandler: So, if your device is just a couple of dollars and you want to track something relatively expensive, the IOT device just becomes a consumable. You take it from the factory, batteries included, it will have an operating life of let’s say two years, and then at the end of its life, you’re just going to dispose of it for recycling. But isn’t that a bit wasteful? The environment is such a huge consideration nowadays.

Joannes Vermorel: Like many things, supply chains consume a lot of consumables, including all this packaging that needs to go around the goods to prevent them from getting broken. These are also consumables that need to be recycled. Consumer electronics, in general, are fairly easy to recycle. Plus, we are literally talking about grams of materials, so it’s very light. Therefore, the environmental impact is very faint just because it’s so small, so light, and if done properly, it can be recycled almost entirely.

Kieran Chandler: So, let’s talk about the data that these microprocessors are producing. Could you discuss how IoT technologies could be implemented into existing ERP systems?

Joannes Vermorel: Indeed, implementing IoT into existing ERP systems is one of the biggest challenges. An IoT fleet can generate a massive amount of data. Each device itself might not generate gigabytes of data, but due to the sheer number of devices, the aggregated data is typically vast. It’s usually one or two orders of magnitude more than the typical transactional data you were historically collecting through an ERP.

This requires different IT strategies for processing, such as the NoSQL movement. NoSQL databases can process a lot more data in a much more scalable way. You typically need some sort of big data capabilities to aggregate and process all this IoT-generated data. It doesn’t naturally fit into your historical architecture of your transactional ERP world. It requires extensive support from the IT department to roll out all the necessary components, typically on the cloud, to support the IoT fleet on the ground.

Kieran Chandler: Are there any companies that have successfully overcome this hurdle to do with data?

Joannes Vermorel: Yes, there is. Amazon, for instance, has been pioneering this space. Jeff Bezos’ long-term thinking has been impressive. Back in 2002, he issued a famous memo pushing Amazon towards a service-oriented architecture. This is a good choice if you need services dedicated to scalable event streaming, much like what you need with an IoT fleet. They also acquired Kiva systems, which essentially makes use of IoT by enabling tracking of hundreds of robots operating in warehouses.

Amazon is even pushing further the boundary of supply chain with the Dash Button. This device, which you can buy from Amazon and stick on your fridge, allows you to order additional quantities of a product with a simple press. They’re doing very aggressive things with IoT nowadays, although there are still very few companies moving in this direction.

Kieran Chandler: It’s an interesting prospect that in the future, instead of doing a supermarket shop, I can just open my fridge, press a few buttons, and have everything delivered to me. Other than data, are there other hurdles blocking the adoption of IoT?

Joannes Vermorel: Yes, after upgrading your IT infrastructure to cope with the extra data, the second biggest challenge is security. IoT devices are fundamentally at risk, much more so than computers that are sitting in a well-protected data center. IoT devices can be hacked. In fact, the largest botnets that wreak havoc over the internet are built from hacked IoT cameras, used by criminals to carry out attacks.

So, for supply chain companies wanting to adopt IoT to upgrade their supply chain, the second largest concern would be to implement defense-in-depth to ensure their fleet of devices doesn’t get hacked and used for destructive purposes in the digital or real world.

Kieran Chandler: The purpose of introducing all these smart devices in the network, it feels a little bit too close to a black mirror storyline. Can you define for our viewers what you mean by a “botnet”?

Joannes Vermorel: A botnet, in simple terms, is when you have access to a large number of computers, say one million, and you can use them to do things on the internet. However, these are usually not good things.

Kieran Chandler: Can you give us some examples? Who is doing such things nowadays?

Joannes Vermorel: This is mostly done by organized crime groups. They can take control of these machines and use them to benefit from their internet connection. For example, they can connect to websites and download homepages. The tricky part is, from the perspective of a website, it’s hard to differentiate between a real person and a machine making these requests.

Kieran Chandler: So it’s a machine that’s downloading the webpage, not the actual person?

Joannes Vermorel: Precisely. If a criminal has access to one million machines, they can all request a webpage at once, creating a denial of service attack. The website gets overloaded and no one else can access it. Then, the criminal group will offer their “protection” services to prevent the issue, much like a ransom.

Kieran Chandler: It sounds like these criminals are exploiting Internet of Things (IoT) devices?

Joannes Vermorel: Correct. IoT devices are prime targets because they often have lax security yet possess internet connectivity. This makes them perfect for forming botnets. However, even more severe attacks are possible.

Kieran Chandler: Before we delve too deep into hacking methods, could you explain how we can secure these devices?

Joannes Vermorel: Securing these devices is a complex problem. Even established companies like Intel have had security flaws in their CPUs. For instance, the Specter and Meltdown flaws that were discovered this year had been present for about two decades undetected. However, there are some basic measures to ensure security.

Kieran Chandler: What are these basic measures?

Joannes Vermorel: Firstly, the source code needs to be audited. People should try to breach the device’s security. If they can’t, that’s a good sign. However, without someone attempting a breach, we can never be certain of the device’s security. Many technology vendors offer bounty programs where they pay individuals who can find and demonstrate security breaches. This encourages the ‘good guys’, the white hat hackers, to make these systems more secure.

Kieran Chandler: You have mentioned incentives for good people to help secure your IoT deployments. It’s a multifaceted issue without one simple solution, but investing in security is definitely a prerequisite. I’m particularly intrigued by the idea of having “good guy hackers.” Now, IoT definitely appears to hold a lot of potential, but there also seem to be many hurdles to overcome. How do you see things unfolding in the near future? Do you believe that these devices will be something we use daily by next year? How do you envision the future of IoT?

Joannes Vermorel: The future is already here in many ways. For instance, consider what Uber does. They are utilizing IoT at scale with human drivers since they don’t have autonomous vehicles yet. In terms of the supply chain, they use IoT trackers, which are the smartphones of their drivers. They don’t even need to pay for the IoT device since the drivers already have them and install the app. Uber tracks the position of every single vehicle, which is an asset part of the Uber network. They have this overarching control system where they can dispatch demand to the closest drivers and incentivize drivers to be present at certain times in certain locations. So, for me, Uber is already like an IoT network.

The same thing is happening with autonomous trucks. Uber opened the first completely autonomous truck line that traveled across the US. So, I believe in the next few years we will see a lot happening around autonomous vehicles and traceability. Traceability isn’t just about preventing counterfeits, it’s also about ensuring the integrity of your products throughout the entire supply chain. IoT gives us the ability to track every container, every pallet, every box as long as what you’re moving has some value. This is especially relevant if you’re tracking things like vaccines, consumer electronics, luxury goods, and so forth. It’s beneficial to ensure that your entire chain is safe, integrity is preserved, and you have complete traceability and real-time update information about everything within your supply chain.

Smart companies, like Amazon, are ahead of the pack, but they will face competition. These companies will leverage real-time insights down to the location of every single pallet to make their supply chain operations smarter. They’ll become more agile, able to change plans midway through a delivery in response to emerging situations. For instance, in aerospace, if there’s an ‘aircraft on ground’ problem, a real-time alert can trigger the redirection of a shipment that’s already in progress. I believe that these developments are on the horizon. The most significant challenges will likely be implementing the changes in IT and making organizational changes to accept these new ways of operating the supply chain.

Kieran Chandler: That does sound promising as long as it’s implemented correctly, and we don’t end up in a dystopian “Black Mirror” scenario anytime soon. Thanks for another fascinating discussion. We’ll be back next week with yet another episode. Until then, we’ll see you soon. Goodbye.