Constellation Insights

Microsoft IoT Central enters public preview: If your enterprise wants to empower line-of-business users to build IoT applications quickly, Microsoft says it has the answer in the form of IoT Central, which is now in public preview. 

IoT Central is a counterpart to Azure IoT Suite, which offers deeper customization capabilities and access to underlying services, with the tradeoff being a need for skilled developers. Both services provide quickstart application templates. IoT Central's browser-based Application Builder environment employs a wizard-like approach for creating models of IoT devices, setting application logic and parameters, and testing via simulation before live deployment.

IoT Central taps into a rising tide of interest in low-code platforms aimed at empowering citizen developers. (Microsoft has some experience in this area already with PowerApps.)

While the focus is on simpler scenarios, IoT Central applications can scale out to millions of devices, Microsoft says. It leverages multiple components of Azure IoT Suite, such as Azure Hub for device connectivity, but runs as a fully managed service. (For a deeper dive into IoT Central's technical details, go here.) Microsoft is offering a 30-day free trial for up to 10 devices; preview pricing for larger deployments is set at $0.50 per device per month.

POV: The speed and simplicity Microsoft is promising IoT Central will deliver could be compelling for many enterprises, particularly ones who have struggled to get IoT applications out quickly with other solutions (including Azure IoT suite). But buying decisions will require a thorough side-by-side assessment of the feature sets for IoT Suite and IoT Central with regard to customization capabilities and pricing. For its part, Microsoft says IoT Central offers a "medium" degree of customizability.

Still to come are important features for IoT Central such as pre-built integrations with enterprise apps. Microsoft says those are coming for Dynamics, Salesforce and other products.

"It's always good to see vendors making things simpler," says Constellation Research VP and principal analyst Holger Mueller. At Build 2016, Microsoft held a workshop for Mueller and other analysts in which they connected a Raspberry Pi to Azure IoT. The task was ultimately manageable but required a lot of steps and had some stressful moments. "It's good to see it simplified even further."

UPS faces holiday crunch—again: This year's holiday shopping season saw an uptick in online sales so strong that UPS had to delay some deliveries by one or two days, the Wall Street Journal reports. In an attempt to keep up with demand, the company has mandated that its delivery drivers work up to 70 hours over an eight-day period. While that's allowed under federal law, the Teamsters union, which represents drivers, is planning demonstrations and potential legal actions if the mandate isn't reversed.

UPS has had these holiday congestion issues before, despite hiring thousands of additional seasonal workers as driver helpers and distribution center staff. It expects to deliver 750 million packages between the U.S. Thanksgiving holiday and December 31, a rise of 5 percent over last year, according to industry publication Freightwaves.

POV: As the world's largest package delivery company, UPS is a crucial bellwether in the rapidly changing world of logistics. It is currently in contract negotiations with the Teamsters, who are working under a five-year pact that expires July 31. The new deal will most likely be for a similar term.

While the bulk of negotiations will concern working hours, wages and benefits, talks could also broach the role of self-driving trucks and other autonomous forms of delivery in UPS's operations. The company has been testing drones for years, including one earlier this year that launches from a truck's roof, delivers a package adjacent to locations along the driver's route, then returns to its docking station.

The drones have huge potential for UPS's bottom line, as they could save many millions in fuel otherwise burned by trucks going to those routes and would also provide obvious operational efficiencies. They also could have a negative effect on the UPS driver fleet's livelihood. The Teamsters have said they are "closely monitoring" the development of drones at UPS and based on their statement, won't be avid proponents of them.

Overall, the next few years will be interesting times at UPS as it juggles the challenges of meeting consumer demand, negotiating with its workforce and managing the rollout of new technologies.

Researchers predict most software will be written by computers in 2040: Scientists at the U.S. Department of Energy's Oak Ridge National Laboratory argue in a newly released paper that by 2040, the bulk of software code will be written by machines, with humans playing a highly diminished role:

The combination of machine learning, artificial intelligence, natural language processing, and code generation technologies will improve in such a way that machines, instead of humans, will write most of their own code by 2040. This poses a number of interesting challenges for scientific research, especially as the hardware on which this Machine Generated Code will run becomes extremely heterogeneous. Indeed, extreme heterogeneity may drive the creation of this technology because it will allow humans to cope with the difficulty of programming different devices efficiently and easily.

The authors cite a variety of projects and technologies that already exist for machine-generated code, including the Defense Advanced Project Agency’s (DARPA) Probabilistic Programming for Advancing Machine Learning and Microsoft's DeepCoder. In the future, "if a human does need to write some code, they may find that they spend more time using autocomplete and code recommendation features than writing new lines on their own," they write.

POV: The paper is worth a read (h/t to the Register) but is framed as speculative, and perhaps rightfully so. Its authors give no specific reason for the 2040 prediction, and barely get into important software development topics such as requirements gathering, testing and security. But the paper's focus on the very real problem of increasing hardware heterogenity, and how machine-generated code could help mitigate it, is a provocative one.