As mentioned in Part One of this interview on Monday, Orange is finding a good new revenue stream in analysing mobile data for, among others, France’s tourism industry.
In the Internet of Things (IoT), Orange Business Services‘ answer to the analytics challenges it foresees for 2016-2025 are in four components. Select; Connect; Manage; and Control. Here, Olivier Ondet tells Jeremy Cowan what customers really want, beyond connectivity.
Olivier Ondet: The four components correspond to four steps that every IoT analytics project has to go through to get to market. Selecting the objects and sensors to generate the raw data, you need to get data from somewhere. Then connecting these objects so that the data is getting to the cloud platform somewhere, today there is not one single IoT network to do that but there are several solutions depending on what you’re looking for. We have all the solutions with different footprints.
The next part is managing the data. It starts with storing and protecting it, visualising it, and then it’s about injecting that into the IoT system of the company or, eventually, the consumers so that it’s used and optimised.
In terms of a pricing structure, at the beginning of the story we are in Euros per object that we select, test, validate, and integrate. So that connecting is usually in Euros per year per object. When it’s about managing it’s also about Euros per year per object.
When it’s about controlling, usually this is an IoT project where we sell a number of man-days of the developers who are doing the coding to make sure the right information is integrated.
It’s a modular offer. We have Orange Assets and partners and a partnership strategy. In Orange Assets we have a catalogue of objects which we already deploy in real life projects, so they are compatible with the whole chain and they can be reused. There are many use cases in different verticals. It’s increasing our catalogue, expanding with each project.
Today there is no one single connectivity solution, it depends on what the B2B customers are using. If it’s coverage then the best solution with the widest coverage is 2G, based on strength in Europe but the US operators never went very far with 2G.
In France, as Orange our coverage ratio of territory with 2G is 98% of the territory, while with 4G we have 80% of the population but 40% of the territory. So if it’s a very rural case the 2G connection may be better. In Belgium, however, which is a flat country, the coverage is about the same for 2G and 4G, so 2G is cheaper than 4G but the coverage benefit is not so high. A lot of IoT is going through old 2G (infrastructure) but in remote areas it’s a challenge.
If people are looking for the quality of service and bandwidth, which is the case for the car industry, the Wi-Fi hotspots on buses and trains and so on, the best solution for that is 4G.
Even in some cases in factories when there’s lots of objects, if the plant is in a remote area then it’s got to be 2G or 3G. There are cases when the B2B customer asking us, “Can I upgrade the site to 4G?” They pay to get 4G on their site so that the factory can become (Industry) 4.0 with all the right connectivity to do the job.
In industrial process we have a solution which is running on cellular SIMS. So things can work on 2G and 4G but initially the SIMS were an Orange France SIM, or a BT or T-Mobile SIM. Now there are SIMS that can be remotely monitored to become the SIM of a given operator in a given country.
That’s typically what we are now selling to the car industry. They design and build the car in China, for instance, but when the car arrives in Hamburg, Germany, the SIM becomes a T Mobile SIM card and it’s going to behave locally as a German card. If the car is sold and it crosses the border into France, then remotely it’s going to become an Orange France card which is going to behave locally as a local French SIM card. That is a way to get a cost competitive bandwidth, but it’s also a way for our B2B industrial customers to have much simpler processes because they only put one SIM. This type of SIM is a bit more expensive obviously, but instead of having pools of SIMs for each country where they are going to ship their goods they can have one single processing system.
IoT Now: What happens when the car is operating along a border area on a regular basis? How do you deal with the roaming issues?
OO: It’s working. It can happen but it’s marginal. You always have these kinds of issues, people are used to that, it’s a bit more expensive and they may have a bit of quality of service problem when they cross the border. They know that but it’s not a major issue.
There are also cases where what our customers are looking for is energy saving; they want the object to last for 16 years instead of three. Particularly if you have to change a water meter.
The business of a water meter is to collect the data. Initially, they were collecting the data every year with people going and looking at the meter, or going with a car and getting the information from around 10 – 20 metres away with radio (links). But with LoRa they can get the information every day, so there can be real-time leak detection and real-time in the water business means that you find a leak in the right day rather than the right year. If you have to change meters every three years then the cost is obviously expensive, especially because you change the whole meter, you can’t only change the battery because there is water everywhere and it would be a nightmare, so you change the whole meter.
‘France is our test lab’
For that reason we have started deploying in France as our test lab, LPWA technology using LoRa, standing for long range. These are LoRa antennas which are put on our traditional cellular sites and they connect the object with an energy saving mode.
We made a choice to deploy a national network but with indoor coverage, starting with 17 urban areas in France. We’re in the process of employing a LoRa gateway wherever the business is, so the motorway station I mentioned was not in our initial plan but in two weeks we just altered the deployment so that the motorway station was covered. I had another case with a parking sensor business case in another part of France that was not in our initial plan which we got covered in one month.
Our idea with LoRa is to test all the LPWA cases so that we and our B2B customers can keep on using LoRa when it’s starting business in factories; these are good areas where LoRa can last quite a long time.
When they go from B2B2C cases there will be a time in two or three years when the 4G technologies will have evolved to also be LPWA-compatible, in the energy saving mode.
Local gateways
Local gateways are typically the case with insurance companies when they say, “I’ve got sensors in homes. I can provide a better insurance service if I can monitor smoke, leaks, doors opening and so on, and see that everything is monitored on a daily or hourly basis. Not only in reimbursing €2000 of damage, it will only be €200 because I will have warned the customer. The customer will be happy to have been warned by the insurance company that there is a problem going on, and they will say that “this company is much better than the others because they’re not only paying but they’re providing a better service.”
In that case we have designed these two local gateways, these are Orange objects actually. One has the form of a cylinder which is about six inches (15cm) wide, it’s a loudspeaker particularly for elderly support in the home, you can press it and have somebody on the phone. Obviously the insurance company is providing the service with the people trying to spot a business case. And it’s a ZigBee / 3G router which connects locally with ZigBee objects inside the home and it’s connecting with 2G/3G so it can work in every home everywhere where 3G is providing coverage.
The partnership strategy we have with connectivity is roaming. We have created a global M2M station with the GMA, with a competitive roaming agreement and processes with T-Mobile and Telia Sonera. Talking about devices, we do have partners which are object manufacturers.
We have a data management platform, with storage, processing, device management, and we have integration skills in France and Belgium and on large international projects with Orange Applications for Business. We have 2,500 developers that deliver projects to our customers, making stuff work M2M.
Each time we have a project we source between what is going to be reusable by our customers, in which case we develop them as part of the product, and what is customer-specific.
Gradually, this is also a way to make a more competitive project because then the customer is not paying for the full feature and they are benefiting from the features previously developed in the roadmap. That’s a typical benefit of cloud.
Here, again, we have a partner strategy because there are some business problems in the world which have already been solved, like indexation, prediction, correlation and business collaboration, urban hypervision for smart cities with a company called Geotech.
We are partnering with them and doing the integration project using their APIs and their platform, not reinventing the wheel just connecting the platforms with each other.
Our data management platform also has APIs that are exposed to be used by local integrators or by our customers’ integrators. We are a reasonably strong integrator in France and Belgium on a large international project, but we are not an integrator in Germany or Spain, for instance. So in these geographies we just have APIs open so that local integrators can do the job of delivering an end to end project.
IoT Now: In data analytics, you mentioned that you’re strong as an integrator in France but working on it in other geographies. Is this going to be the greatest area of revenue generation for Orange in IoT for the next few years?
OO: All four components have good potential for revenue generation. Then it depends on the geographies and business cases. Particularly in connection worldwide, in all 29 geographies. The purple part (see top slide. Ed) is equipment research, so typically smart home business for this kind of vertical or even smart cities is especially important, not in industry where they already have the sensors. It’s an equipment resale business but with these objects there is value in understanding how they work, testing them end-to-end and making them.
In the smartphone business there is a duopoly with Apple, Samsung and a bit of Microsoft, and their objects are highly complex but there are not so many of them. In the IoT space there are many different objects of different types so people’s technology is specific to them. Integration is indeed a business which is much more local. Yes, there may be growth there but we cannot do that organically.
IoT Now: You seem to see a bigger opportunity not in the consumer Internet of Things, i.e. connected domestic devices, but more immediately in industry and manufacturing. Is that fair?
OO: Yes, there is a lot of potential in industry but we also see many smart cities projects popping up. There was a customer case in Doha, Qatar. Automotive also has a strong impact but also health for B2B business.
IoT Now: Presumably you’re working through the channel?
OO: It’s B2B companies; in automotive it’s the car manufacturers, it’s car rentals, it’s insurance companies, it’s every kind of company that has a fleet of vehicles, these are customers that we have and we are making significant growth with connectivity and services.
I mentioned four specific areas, smart cities, industry, automotive and health. They are all B2B customers, B2B2C. Industry is pure B2B, but smart cities, automotive, health and industry — if they also have consumer goods in the manufacturing area — then in that case they are going to be B2B2C.
But yes, within the Orange Group B2B and B2B2C is a strong part of our ambition in IoT. Then our colleagues from the B2C business, Orange has over 240 million customers, and most of them are a number of B2C customers, consumers, and my colleagues from the B2C business also have an IoT strategy which is mostly about smartphones and bringing daily services to make everybody’s life smarter.
We are coming a lot from the industry business but we do have a lot of potential. So it’s a bit more contrasted than what you said, but historically you’re right.
IoT Now: Okay. When you’re working in areas and industries that didn’t even exist five years ago, how do you search for the skills and people in data analysis that you need?
OO: That’s a really good question. We have a modular offer, with the Devices, the Network, the Data Management, Control and Integration skills.
We’ve created four strategic verticals where we have teams whose job is to completely understand the value chain of these four verticals. These are B2B2B2B value chains where the big question is, “Which B should I work with?” and then get the use cases. Then we can trigger the right skills to mobilise them on the project which is something we are used to doing as an IT company.
So these four specific areas are the four I mentioned, smart cities or territories, automotive, healthcare and industry. We do have a team on smart cities, a team of 40 people and in healthcare a team of 30 people, they are specialists in these areas and are business developers. They mobilise the rest of the operational structure of the team to get the components they need, but also to make the offer evolve to match their requirements and, more globally, so it can be used in other verticals.
Taking the example of LoRa for business, LPWA, it is intensively driven by our smart cities guys, because that’s where you see a lot of sensors today, but it does have positive effects for industry because it’s a way to connect objects that were not connectable before in industry. Before there were economies of scales triggering the launch of a network for smart cities. So that’s the case for the network.
There is a similar story with algorithms. We have a team of data scientists both designing the products and doing projects, these teams are working on every kind of project in every vertical, but they are gradually becoming more and more expert. Today, for instance, we have real experts in the tourism industry in data science because, with analytics of the mobile network, we have developed good skills in understanding tourism from using the data from the mobile network. So we are bringing a lot of value to the tourism industry, which was co-created with the tourism industry. Now we have created some expertise which we can reuse in a vertical way.
Olivier Ondet, vice-president of IoT and analytics at Orange Business Services was talking to Jeremy Cowan.
Comment on this article below or via Twitter: @IoTNow_ OR @jcIoTnow
