For the last couple of years, Salesforce has been making moves in the Internet of Things (IoT) market by activating a variety of functionality that allows the company’s cloud-based infrastructure to collect, organize, and custom develop software that taps into data and functionality from the “things" — sensors, devices, etc. — that are a part of the Internet of Things. Looking to tie all of that functionality together into a single pane of glass that not only sets rules and illustrates the connections and flows between such devices, but that also allows them to be monitored and interrogated for their status and data, Salesforce has announced IoT Explorer.
ProgrammableWeb was briefed on the announcement and given a demo by the company’s executive vice president and general manager of IoT Woodson Martin and director of IoT product management George Saines. The video and audio of that briefing and demo are embedded below, followed by a full transcript. All of ProgrammableWeb’s videos are also viewable on our YouTube channel. You can also tune into the audio part of our videos and any audio-only podcasts through our Soundcloud channel (where you can subscribe with your podcatcher as well).
One of the best take-aways from the video that should not be lost on viewers, listeners, and readers is how the Internet of Things makes it possible for some very old-school businesses that might never have dreamed of turning themselves into digital platforms to do exactly that. A good example of this is how GE uses its Industrial IoT platform Predix to turn the aircraft engines it manufactures into a digital platform that enables the company to offer significantly better service to airlines than its competitors. The engines come equipped with thousands of sensors each of which stream data back to GE where artificial intelligence can ingest the data and anticipate a failure in a way that humans never could.
When companies transform previously non-digital offerings into platforms in this way (the art of “platformization”), application programming interfaces (APIs) are one of the key enablers because there’s really no such thing as a thing on the Internet of Things that doesn’t have an API as its primary interface. This way, other software and developers can gain access to its functionality and data. That API could be on the device itself, on a local gateway that acts like an API proxy, or it could be up in the cloud (the way the API for every Fitbit is up in the cloud). But however that solution is architected, the so-called “things” on the IoT invariably have APIs associated with them.
If GE sounds a little out of your league as a comparison point, then consider the example that Martin brought up where similar sensors (perhaps not nearly as many) are in the trash barrel that you drag out to your curb once a week. “[Rehrig Pacific doesn’t] make trucks, they don’t collect garbage,” said Martin. “But they make the plastic injected molded trash cans that you wheel out to the curb and they’ve embedded RFIDs into these trash cans and they’ve put these sensors on the trucks so that when the trucks drive around on the route, they can detect which trash can is which.”
There are many use cases once a “technology company” like Rehrig Pacific — one that a decade ago probably never envisioned itself as a digital platform — is able to amass and correlate such IoT data on behalf of its customers (the trash pickup companies) and, in turn, their residential customers. For example, Martin alludes to a use case where, by correlating truck geolocation data and barrel pickup data, it would be relatively easy to not only spot if a trash barrel has been stolen, but potentially who stole it.
But you could also imagine all sorts of other interesting business intelligence that could be mined from the sea of data that’s collected along with the garbage. For example, to drive the cost out of garbage collection routes, trucks probably have to make the shortest stops possible; ones where the mechanical arm lifts the barrel and dumps it into the truck and the driver moves on.
But you and I have all seen those families that constantly overload their barrels in a way that prevents the truck from making an optimal pickup. Maybe one of the two people hanging off the back of the truck must deal with those barrels manually. That stop takes a few extra seconds, burns a little more fuel, and ultimately costs a little more money. One stop? No big deal. But a hundred or two hundred stops like that and now, it’s cutting into the trash collector’s bottom line.
However, when equipped with the right data — the barrel’s RFID tag number, the location and timestamp of the pickup, maybe even a photo from a truck mounted camera — the trash collector can spot, identify, and communicate with the worst offenders in an effort to optimize route operations. Maybe the customer should be charged more. Or maybe the customer should be given an extra barrel (or a larger one that’s compatible with the collector's trucks).
But why Salesforce; traditionally more of a CRM company?
As Martin explains, “It’s really the marriage of the customer record” said Martin. “Everything you know about a [customer] in Salesforce; what trash can are they supposed to have at that location and the data that is streaming from the Internet of Things at the same time. You put those two things together and then you use logic, business rules, artificial intelligence over time to help make all those customer facing business processes smarter.”
In addition to offering a single pane of glass that Salesforce IoT users can now use to get a more comprehensive view of their networks of things and what those things are doing, the offering also looks to make it relatively simple for non-programmers to “wire-up” their use cases and business rules. According to George Saines, "we've built a UI that enables business users, not software engineers, to go in and create these rules. These rules are just about as simple as are shown here in these quick descriptions (referring to the demo at about the 15:00 mark).” Based on its exposure to many vertical industries, Salesforce will also be offering pre-built rulesets and use cases on a per industry basis. For example, the waste management example above is obviously one that could be templated and re-used all over the world. In the demo, Saines shows a use case where solar panels are the "things" that are connected to the Internet of Things of the fictitious company Cirrus Solar. As the panels stream data back to Cirrus, Cirrus can remotely diagnose any issues with their ability to produce power and potentially fix those issues without "rolling" a truck to the customer's installation (the screenshot below was taken from the demo).
To hear what both Martin and Saines had to say and demonstrate, watch the video below or listen to the audio-only version (also embedded below). A full-text transcript of the interview appears further below.
Video Interview of Woodson Martin and George Saines About Salesforce's IoT Explorer
Editor's Note: This and other original video content (interviews, demos, etc.) from ProgrammableWeb can also be found on ProgrammableWeb's YouTube Channel.
Audio-Only Version of Interview with Woodson Martin and George Saines About Salesforce's IoT Explorer
Editor's note: ProgrammableWeb has started a podcast called ProgrammableWeb Radio. To subscribe to the podcast with an iPhone, go to ProgrammableWeb's iTunes channel. To subscribe via Google Play Music, go to our Google Play Music channel. Or point your podcatcher to our SoundCloud RSS feed or tune into our station on SoundCloud.
Full Transcript of David Berlind's Interview with Woodson Martin and George Saines
David Berlind: Hi, I'm David Berlind, Editor in Chief of ProgrammableWeb. Today is Tuesday, October 10th, 2017 and we're here today to talk to Salesforce about their announcement in the space of the Internet of Things. Joining me right now is Woodson Martin. He's the evp and general manager of IoT for Salesforce. Woodson, thanks for joining us.
Woodson Martin: Hey, David, thanks for having me. Great to be here.
David: Salesforce, obviously a long and storied history in the Cloud and all that. As of late, in the last couple of years with the Internet of Things becoming a rather popular subject matter, seems as though Salesforce could not resist entry into that market as well as a bunch of other ones that have come up along the way. Why don't you give us like the two minute history of where Salesforce has been with the IoT? Then we'll dive right into the announcement.
Woodson: Great, David, thanks. Salesforce first got into IoT in 2015 and since that time [00:01:00] we've been working with some amazing, fascinating companies, frankly, across a wide variety of industries. Companies like Schneider Electric, KUKA Robotics, Ecolab, these are companies who are focused on using IoT as a way to rethink business models and customer experiences in their industries.
We got into it in the first place because one of the things customers were telling us was while they were investing tons of money in stuff like these sensor devices — the technology infrastructure to make connected products in the first place — it was hard to focus on, to factor in the customer experience for the person behind the product. Of course, that customer behind the product has always been the focus for Salesforce. It's really a natural fit to bring the way you think [00:02:00] about CRM, the way you think about how you service your customers, how you engage your customers, how you sell to your customers, in your strategy for IoT.
That's really where we've been focused for the last couple years. We're excited today to be making an announcement of a new product in the category. One that brings all of that learning and makes it simple and easy for people who use Salesforce today to tap into the Internet of Things to start to use that data to drive, to build new customer experiences, to iterate on those and to really differentiate their products and services.
David: When you talk about the customer behind the product, you sort of see this as an extension of customer relationship management. We're talking about customers that use something that may be generating some amount of data and you want to tap into that data stream to enhance the record that some organization might have of that customer?
Woodson: Yeah, so that's one way to think about it. Certainly that's a pretty simple way to think about [00:03:00] it. Let's make it real by talking about an actual customer, maybe, and sort of the kind of things that we're seeing people do. I'll talk about a company called Rehrig Pacific. You've probably never heard of them but you probably have a truck like this one that drives in front of your driveway and picks up your trash cans every week.
David: I do and it makes a lot of noise when it comes by. I'm always on a call, too, when that happens.
Woodson: For me, it happens at like 5:00 in the morning so I'm usually trying to sleep. The interesting thing is Rehrig Pacific; they don't make trucks. They don't collect garbage but they make the plastic injection molded trash cans that you wheel out to the curb. They've embedded RFID's in these trash cans. They've put these sensors on the trucks so that when the trucks drive around the route, they can detect which trash can is which.
That's important because it turns out that not every trash client is as, [00:04:00] let's say, honest as I am or you are. Sometimes people steal the trash cans and they take the big trash can that costs more money and they stick it front of their driveway to get rid of all their trash. Now, here's Rehrig Pacific, a company that has long been in the game of plastic injection molding, is suddenly becoming a data provider to the waste management industry and helping them to monetize all kinds of things they couldn't before. Helping them to provide a better and more proactive service to their customer.
If somebody steals your trash can, you get it back, they know it's gone. They can provide a better service there. Well beyond that, revenue recovery opportunities when people are using the wrong trash cans that aren't the ones they're billed for. All kinds of stuff. It's really the marriage of the customer record. Everything you know about a company in Salesforce, what trash can are they supposed to have at that location [00:05:00] and the data that actually is streaming from the Internet of Things at the same time. You put those two things together and then you use logic, business rules, artificial intelligence over time to help make all those customer-facing business processes smarter. That's the fundamental idea behind Salesforce IoT.
David: You guys do have also a play in the artificial intelligence space, as well, so I'm assuming that some of that technology can be brought to bear to analyze the data?
Woodson: Absolutely, you know, you think about what's ... there's been IoT and artificial intelligence used in combination for a long time. Most of the artificial intelligence and IoT today is doing something called anomaly detection. It's basically looking across patterns of large amounts of data to try and get better at detecting when there's a problem or there might soon be a problem with let's say a piece of equipment or a device.
That's super valuable, of course. There's a lot of prior art [00:06:00] there. What's really newer, when you bring together the world of IoT and you bring together the world of CRM, suddenly you can use artificial intelligence to reason across all of that. A lot of the value for a company in figuring out what to do when an anomaly is detected is on the customer end of things. What works well in the customer experience?
Let's say there's a problem and we think the machine might fail in a month. What if the contract expires in a month? Is it better to go ahead and fix the customer's problem now while it's still under warranty so they feel better about the renewal a month from now? Or is it better to just wait, let it fail and have the customer have an important priority around getting it fixed? I don't know. It depends on your business. You can now reason over all that data, the chain of data. It's not just detecting the anomaly in the first place but kind of the responses you take, what you do and what the impact [00:07:00] of those things is over time with customers you can optimize that process.
David: Woodson, you've made several investments over the last couple years in the IoT space. Sounds like it's working for you and some of your customers. What's the big announcement today?
Woodson: Today, we're talking about IoT Explorer. We've taken everything that we've learned over the last few years in IoT and baked the functionality directly into the core Salesforce platform to tap into these data streams that are coming largely from IoT device management platforms. You might think about systems like AWS IoT or Azure IoT or even in the industrial world platforms like GE Predix or Honeywell. Using data from those platforms to orchestrate in a simple, low code way these customer experiences for your end customers.
To bring to bear to that, all the data [00:08:00] that you have in Salesforce about those customer records, the customer history and then to activate using any part of the Salesforce platform, any Salesforce cloud to take action. Whether through human channels, by activating your sales team to pursue an opportunity you detect, through device data or your service team to go out and proactively fix a problem for a customer. Even to drive the way that you communicate and interact with customers with digital channels like email or mobile messaging and packaging that all up in a very simple way that helps our customers get started with their IoT strategies and really iterate in this customer experience.
David: How is that different from what you had before because that's the sort of thing that it sounds like you need anyway if you had access to all the data that you've been providing now for a couple years.
Woodson: Yeah, well what's really cool about this is we have a lot of these ideas we've been working on for a while. We've now put them all into a simple package that we're making very easy for all of our existing customers to take [00:09:00] advantage of baked deeply into the core Salesforce platform, taking advantage natively out of the box. No more need to do data preparation, to move data that you may be managing in Salesforce into a separate environment for your IoT processing. It's all done right in the context of where you're managing your data today and fully leveraging the Salesforce platform. It's a great step forward, simplifying the world of IoT for our customers.
David: The proverbial single pane of glass, as they often call it?
Woodson: Yeah, sure.
David: I know you guys are prepared to show a demo of this today, so why don't we take a look.
Woodson: Yeah, my buddy, George Saines, is here and he's going to walk us through a brief demo of IoT Explorer.
David: We'll move over to George Saines. He's the Director of Product Management at Salesforce. George, why don't you go ahead and give us a show of what this IoT Explorer looks like.
George Saines: So, as [00:010:00] Woodson was just mentioning. This is all about enabling the customer to solve business problems with device data. The demo I'm going to show you today is centered around a use case. We've got a fictional company, Cirrus Solar. They make and manufacture connected solar panels. Now, Cirrus Solar, has three problems at the moment. The first is that their cost of service is too high. The second is that their sales pipeline value is too low. The third is that their case origin or the way in which they're interacting with their customers is primarily reactive as opposed to proactive.
Cirrus Solar had all this device telemetry yet they continue to interact with their customers via email and phone. Now Cirrus Solar makes hundreds of thousands of these panels but we're going to talk to day about one of their customers, Newton Farms, which is a sustainable agricultural business located in Livermore, California. Newton Farms has a brand promise to their customers to produce all the food using sustainable, off the grid power. The relationship between Cirrus Solar and Newton Farms is really important. [00:011:00] The way that Cirrus Solar actually comes in and starts to address these three problem is the first step, they just log into the Salesforce platform. They do that, they go to the set-up tree and they type IoT. In IoT, you've got all the objects that are necessary to build out a complete use case in the app. I just want to emphasize this because this is different, this is new and this is really exciting for our customers.
They're going to click orchestrations and they're going over to what you think of as fleet view. Cirrus Solar, manufacturing these hundreds of thousands of panels, wants to govern what happens when a panel is in a particular state. Let's say a panel is generating no power. When panels are installed, they are installed in tracks and a track could potentially have a problem where it's simply not generating any power. Cirrus Solar knows from past experience that the easiest way to do this, to resolve this for the customer is to call out to the inverter control plane and power cycle the unit. Cirrus Solar doesn't even really get involved. The customer doesn't even know that there was a problem. [00:012:00] It's possible that a track could be generating just less power than its rated to. In this case, what we want to do is actually reach out proactively and get the customer engaged so that they can self-resolve the problem. Cirrus Solar knows from prior history that this is the sort of thing that a truck roll would be inappropriate to solve both from a cost and an expediency perspective.
If there is actually a problem with the panel, though, we do want to get a truck out on site. Up to now, we've been talking about [Salesforce] Service Cloud. This is really cool because we can get engaged and deeply integrate with what we call Field Service Lightening, which is a mobile technician focused application that helps truck rolls be more efficient.
Finally, at the beginning of this demo, we had talked about these two predominant problems. One was the service cost and then the second is that they're not generating enough sales opportunity value. This state, the purchasing capacity state, looks at all the device telemetry coming from all of Newton Farms' solar panels. If it's getting close to where they're actually having to draw power off of the traditional grid, what they can do is we can open a sales opportunity for [00:013:00] the Sears Solar sales team to go out and sell more panels.
We've been looking at the Fleet view. This is all the panels. Let's take a look at what happens to one panel, in particular, just to show how this all works. What I'm going to do is I'm actually going to simulate a no power generation event. When we do, we're going to go from the panel monitoring state over to the no power generation state. You'll see we just moved over. We're tracking a few panels here. If we go over to our Service Cloud and we refresh. If I can actually get my screen here. We'll see that we've actually just created a case in Service Cloud to let the service team at Cirrus Solar know that the issue was self-resolved.
If we go over to the low power generation state, I'm going to simulate a low power signal occurring here. What we're going to do is go over to my mobile phone here. We'll see that the IoT Bot just mentioned me. [00:014:00] I'm Jeff Manough, I'm running my business. I've very busy, but I notice that Salesforce IoT has my back here. All I need to do is click this link and I'm going to be linked over to a community portal FAQ where I can see all this information, all this content that would help me resolve this problem. I don't need to reach out to Cirrus. I don't need to send them an email. I don't need to get them on the phone. I can just go and fix this immediately.
David: OK, George, I'll do a question for you here, which is I'm looking at a fairly sophisticated user interface. These spheres that represent different assets and different states of the technology of the power grid, so to say. You pulled up some rules. I saw those. How difficult is it to create all that? Is that something that just comes out of the box or did somebody have to actually do a little bit of coding to put those rules in and to create these spheres an interconnect them in a way that closely mirrors the [00:015:00] process at the customer site?
George:That's actually a really good question. The answer is that we've built a UI that enables business users, not software engineers, to go in and create these rules. These rules are just about as simple as are shown here in these quick descriptions. What we do is you click over to the rules. To your point, some of these are going to come right out of the box. We're going to be having industry specific packages that can be installed from the app exchange that will enable you to just get running 0 to 60 in minutes. For those —
David: By industry specific, you mean like there are different verticals and rules are pretty common to some of those verticals. Take medical, manufacturing or agriculture and the rules generally seem to be the same so you pre-bundle some of the more common ones?
George:Not just the rules but actually the states in the entire use cases are often times very common. Where I was going with that, the next step is that should that package not quite be what you [00:016:00] need, the process of going, modifying, and iterating and then up here in this button activating or putting into production these use cases is the sort of the thing that we can see our trailblazing customers doing in days not weeks, not months. This is really simple, really easy stuff.
David: I think that you're covering a lot of the different possibilities. What about the part of connecting the sensors that are the things, in the Internet of Things, to the system? How difficult is that?
George:What we do is we expose ... it's actually really simple. We use the same platform endpoints that every other platform service provided by Salesforce uses. Anyone who is familiar with any part of the platform, will feel right at home sending device data to Salesforce IoT.
Woodson: I was just going to chime in and say we're also working very closely with a lot of the IoT platform players. We talked about AWS IoT earlier today [00:017:00], to provide package integration between their IoT platforms and an IoT Explorer so it's really simple and right out of the box for customers. Starting November 6th in San Francisco, we'll be hosting our annual Dreamforce Conference where customers and partners, 100,000 people or more out at Dreamforce to basically get their hands on the technology, to understand how the technologies fit together to help them solve business problems. In IoT specifically, we'll have 17 different customer case studies on show. We have built Lego City that emulates the city of San Francisco with a whole host of smart-city technologies simulating the use cases and how these flow through from actions that happen with physical devices [00:018:00] that spawn services that include rolling physical trucks on the floor. It's an amazing visualization. It helps people understand, not only of course, these real customer use stories, but also illustrating a whole host of them right here in one physical space with the IoT Lego City. It's going to be an amazing opportunity to learn more about IoT and bringing together the power of CRM with IoT platforms brings that amazing ability to get proactive in sales, in service, and in customer experience for every connected device.
David: If you want to go to Dreamforce where do people go to find out more about that?
Woodson: Dreamforce.com is the shortcut. Check it out, we'd love to see you there.
David: For the IoT Explorer, where can people find out more information about that?
Woodson: On our website, Salesforce.com/iot.
David: Perfect, that's really easy to remember. Woodson Martin, evp and general manager for IoT the Internet of Things at Salesforce. Thank you [00:019:00] very much for joining me. Thanks very much to George Saines, the Director of Product Management there, great to have you today on the show.
Woodson: Thanks, David.
David: I'm David Berlind, Editor in Chief of ProgrammableWeb, bringing you another one of our video demonstrations on the ProgrammableWeb podcast. For more of these you can go to our YouTube channel. We have one up there. We've got a couple other of these videos loaded up. We'll have more coming. You can also come to ProgrammableWeb and find the article embedded in my write-up of this particular story along with a full transcript of what Woodson and George said. Thanks very much for joining us.