March 20, 2018
Autonomous cars have been a hot topic for a couple of years now. It is indeed a very interesting and complex technical challenge connected to potential resolution of very serious problems and innovative use cases. But it is also about culture and liability which I believe always takes very long time to cope with.
The tragic news that a car in an autonomous trial in US killed a woman yesterday can’t be a surprise to anyone. There has been a lot of accidents with autonomous cars before, traffic is very dangerous and cars kill people every day. Each accident is followed up carefully to understand what happened, why and who is to be blamed for what. Traffic rules, law, policies, insurances, vendor responsibilities etc are in place to help deal with these tragic events. Here is the core problem with autonomous cars – when a car make a mistake, who is liable? And, do I want to meet self-driving cars on the road or fly airplanes without pilots?
Already in the Reuter article we start to see arguments about who owned the software in the car, which brand it had, if the car actually did anything wrong or if a person driving would have been able to avoid the collision. Also that the woman seems to have walked outside the crosswalk, that she had a bicycle and that she might have been homeless. But truth is that it is unclear who is responsible and for what.
I remain convinced that we will not have self-driving cars on normal roads together with other traffic and in normal speed until you and I are ready to have our kids walking to school meeting those cars. My best guess is not before 2030. And given that “rain, snow and ice are particularly challenging for autonomous cars” maybe we just should forget about it in Sweden.
July 7, 2017
Digitalisation is frequently used as the name of the game today. But digitalisation, i.e. conversation from analog to digital representation of information, started some 60 years ago and has been going on ever since. When a process is fully digitalised, from end to end, the process will be much more efficient than before. But most processes today are only partially digitalised which is like building a bridge and leve some meters here and there. Let’s take the example of a smart electricity meter, where the collection of usage data is digitalised and hopefully the receiving systems at the energy company. But then the processed details often are put on paper and sent to the customer who hopefully pays in a digital fashion or god forbid calls customer service.
A fully digitalised process is quite efficient but more importantly it is a great starting point for artificial intelligence in any form or shape to be applied. Imagine that we could move the simplest 30% of the decisions in a process from people to computers. What would that mean to your organisation’s cost, quality and speed of execution, and not to forget your competitiveness? What if we could add an algorithm making our system self-learning thus provide qualified decision support for the remaining decisions made by your staff and maybe even advise to your customers. This is what Digital Transformation is about and that is what we should talk about rather than digitalisation.
It is IoT that adds the last meters of the digital bridge. By connecting the real world to the Internet where we already have people and organisations we can start complete the digitalisation of our processes – a sensor that tells when a door is open, where a car is located, when it’s time to repair a fridge or when the level of carbon dioxide is too high. And this is exactly what the more progressive organisations are working on right now. And since organisations and processes are quite unique there are a huge amount of opportunities for companies to develop algorithms to support specific processes. I already have many members in my alliance for Swedish IoT startups working with machine learning including Ekkono, Watty, Aifloo, CombiQ, BellPal and Imagimob.
The effects of Digital Transformation will be massive in all industries and large-scale redundancy will definitely hit white-collar employees as organisations get their act together. I miss awareness and a serious debate about this in Sweden since it limits our chances to come out of this transition strong, the longer we wait to address the challenge.
March 27, 2017
It’s ten years since iPhone came to market and created a new world of smartphones and apps. It feels like they have been and will be here forever. Given the pace of tech development it is time for the next big thing but it is hard to see any candidates. Glasses and wristbands are accessories and not even close. VR and AR are interesting but probably not what we are looking for. When we look back at 2017 we will see that the next big thing was internet again, this time with things added to it!
When internet surfaced last time we had little or no idea of the impact it would make to countries, businesses and people around the world. We went through the three phases: connectivity, operational value and strategic value
in maybe ten years and the impact has been in the ball park of electricity. Now when we are adding things to the people and organisations already at internet we will go through exactly the same thing again. Connectivity (remember the 50 billion devices) is soon behind us now when we finally get the missing low power infrastructure for IoT, LPWAN
. Since hardware is more difficult to connect than people the first phase will take quite a lot of time, but the two following ones – operational and strategic value – will be much faster since everything needed already is available on internet. We already see companies leveraging things connected to internet today – look at the four nominees to the IoT Enterprise of the Year award in Sweden: Verisure, Volvo Cars, Husqvarna and Assa Abloy
– and this is about to explode across industries and countries. The core deliverables from things connected to internet are safety, sustainability and efficiency i.e. what you will find in most priority lists of today.
In essence: if you want to understand how IoT will develop, try remember how internet happened last time. If you want to understand the impact of IoT, take a sober look at what internet has done to you, your industry, your society and the world. If you want to keep your job and save your company, get going. Internet is the next big thing, again, and we are already in the middle of the revolution. This time ignorance will be no acceptable excuse.
March 7, 2017
I dived into what I called narrowband networks for IoT 2011 since there was an obvious need for an infrastructure that didn’t exist. I used to say “think cat, bicycle and smoke detector” to put the finger on the need. Or translated into requirements: low cost and often small and light communication modules, low cost connectivity and very low energy consumptions. The answer to that would be inexpensive infrastructure which is providing primarily heart beats, events and when needed position. The first solution I found was Sigfox and after a training in Toulouse I wrote this post.
There are four key alternatives for LPWAN emerging: Sigfox, LoRa, 6LoWPAN and NB IoT. Each have it’s strengths and weaknesses and it’s likely they all will play a significant role in the rapidly emerging new generation of Internet.
Sigfox is a proprietary solution with national networks built with Sigfox base stations operated by national Sigfox operators using the same back-end system operated by Sigfox. Sigfox has networks in some 30 countries today, with over 8 million connected devices and have a well developed eco system of developers and vendors on the terminal side.
LoRaWAN is a global LPWAN specification created by the LoRa Alliance to drive a single standard for seamless interoperability across the industrys. Networks are built with LoRa Gateways and LoRa Network Servers available from several different vendors. LoRa alliance has some 400 members and continue develop the standard. LoRa could be compared with WiFi since networks can easily be built with one base station covering an area like a campus or square and expanded to cover cities and countries. But there is no common back-end arrangement for roaming, service level agreements etc.
6LoWPAN is a concept originated from the idea that IP, the Internet Protocol, could and should be applied even to the smallest devices, and that low-power devices with limited processing capabilities should be able to participate in the Internet of Things. It’s worked on by an IETF working group. 6LoWPAN is similar to LoRa since networks typically are deployed independently in a campus, city or building with no global back-end service.
NB-IoT is the narrowband bet by the mobile operators. It differs since it is using private spectrum and typically the infrastructure already in place in mobile networks. The initiative has been developed in record time and most mobile operators are planning roll-outs. The first NB IoT networks are expected mid 2017 and we have just started to see modules, developer tools etc. It is also interesting to follow how NB IoT relate to 5G since support for narrowband has become a key part of what is targeted.
Due to the very big interest and potential for narrowband networks there are several other vendors and efforts to address this need.
LPWAN is now the hottest of questions in IoT
since manufacturers, vendors, users, developers, consultants, operators, teachers and journalists all need to understand what it is, the different solutions available, how to develop applications for LPWAN and when to choose which solution. This is why I try to run as many “Get On Top Of LPWAN events
” as I can to ensure Swedish IoT
will remain leading edge.
Finally, I’m glad to let you know that we finally have a Sigfox operator also in Sweden! IoT Sweden
was just announced and have started to roll out their network in larger cities already. They run their launch event at THINGS March 21
if you’re interested.
December 10, 2016
Technology driven innovation continues to challenge and change our world. Internet is the single most important enabler and the last big thing is the smart phone, which gave us new ways of using the Internet including apps. The iPhone is about 10 years old now and it’s time for the next big thing: connecting things to Internet, where we already have people, organisations and services. This will once again change lives, organisations, industries, companies, cities and governments in the same way and magnitude that the arrival of Internet made last time. But this time it will happen much faster since most of the bits and pieces already are in place. We call this IoT right now but soon it will just be Internet again.
I always think that everything is a science. Regardless of what you look into in more details, it proves to be very complicated. I love the massive opportunities Internet brought and believe these will be even bigger now when we add things to it. But Internet has created a trend which I dislike: by leaving out the physical part of the equation (like running a taxi business without taxis or hotel business without hotels) we are dependent on others doing the hard “physical” part which is connected to places, people, boarders, re-cycling, communities, manufacturing etc. Not that there is anything wrong with entrepreneurs executing on these opportunities, not at all, but we all have to think a little further before celebrating or investing in these initiatives. I see a whole range of challenges, including the vulnerability of these businesses when it comes to policymakers and not the least when the incentives to provide the underlying activities or assets disappear. The Internet part alone also tends to become global leaving only one or few companies in the market (Facebook, Twitter, Google…). Put differently, we simply have to think about the type of society we want to live in.
With Internet of Things this becomes quite obvious. Data is the gold of IoT and a lot of companies are focusing on building business on that data. But without anyone connecting the things, the T in IoT, there will be no data. And when companies make the effort, I don’t think it is obvious that they will share all their gold with anyone anyway. At least not for free. On top of that I’m convinced human beings as well as organisations will become much more cautious with their data and how it is shared and used. Security and privacy are the two big challenges to IoT!
Another dimension of this is that the opportunity to disrupt using Internet has made a lot of people completely lose the respect for the underlying assets or activities. A good example of that is all IT/Internet companies announcing plans to build a car. There is nothing wrong with that, if you have funds to buy all skills and other assets it takes. But I have still to see a car manufacturer announcing that they will start make fridges, shoes, smartphones or Internet search engines. Maybe traditional companies have more respect for the complexity of other industries.
I believe we all would benefit from a little bit more respect for how difficult it actually is to do different things. A hackathon with youngsters to innovate healthcare over a weekend is great education and fun but makes no sense from a healthcare point of view. An app to keep track on parking spaces is easy to do compared to building the infrastructure to provide all the information needed. The how of IoT is about collaboration and eco-systems, where all relevant players have an important role bringing their experience and know-how to the solution.
November 13, 2016
All IoT solutions span at least three industries – collection of data (sensors, gateways, datacom, telecom, etc), managing data (cleaning, matching, analysing, combining, etc), distributing information (IoT value is created when a piece of wanted data is delivered to the right place at the right time, i.e. apps, signs, ERP systems, warning lamps, etc). In the early days of IoT clever people were able to put together all these things to solve a specific problem for a customer or even a number of customers in similar situation. The problem is that any single piece in an IoT solution is quite complicated, so in order to make a really good solution all bits and pieces need to be top-notch. If you need a CO2 sensor you will have to turn to someone who offers the right functionality, quality and price for you solution, at any given time. If you need to have the wanted information from your solution delivered in an app, you need to provide your customer with a top notch app with great UX at any given time. If not your entire solution will look bad in the eyes of the users, even if it’s actually the best one in the market.
The IoT market develops very fast and complete solutions from one vendor, often with a couple of years success behind, are now meeting stiff competition from solutions created by several companies in tight collaboration. These ecosystems are collaborating to provide the best possible solution to industries, applications and customer segments and like if it wasn’t bad enough for the “early stars”, these ecosystems are adapting much faster to customer needs, technical development, legal requirements, policies and trends due to their combined resource.
I don’t believe single companies, industry groups, alliances or standard bodies will determine how IoT will be deployed in different markets or applications. I believe successful ecosystems will. Ecosystems is simply the new how in IoT.