At times this trip feels like a game of Where are Simon and Mary this week; we’ll clock up nine states by the time we get back (including a detour into Ohio via Indiana for personal reasons and refuelling stops in Texas and Colorado). We’ve been relying on the Bluetooth GPS with CoPilot on Windows Mobile to get us to places since we arrived and it’s seen us through California, Nevada, Arizona, Utah and back (in Florida I switched to the BlackBerry 8800). Wherever we go, we want to find important things - the hotel, the airport, the car rental depot, the conference centre, the closest place to get a coffee - and fun things - the nearest mini-golf, somewhere for dinner, a friend’s new flat, the nearest cinema showing Shrek 3. CoPilot is good at that but the maps of the US we have are a couple of years old and not everything we want shows up; tonight we wanted a pet store and a WalMart and I ended up looking the addresses up on Windows Live Search on my smartphone then typing those into CoPilot.
This week we’re in San Jose for the Where 2.0 conference. Google and Microsoft have been announcing 3D map views and we’ve seen a lot of tools for mashups and showing data on a map, until pushpins begin to resemble cartographic measles. But people have also been talking about getting location information where you really need it - when you’re away from your laptop and your DSL connection.
A couple of years ago Garmin bought up MotionBased - a service for logging the routes you walk, bike, paddle or even lawn mow from your Garmin GPS to a Web site called TrailNetwork. At Where 2.0 Garmin announced it has taken the MotionBased Agent software and rewritten it as Communicator, a browser plugin that makes it easy to send routes and locations to a GPS device too. Web sites can embed the code to send an address to a GPS in a few lines of JavaScript; Geocaching.com has already done it and given that the most popular page on the Dunkin Donuts site is the one telling you where your nearest doughnut fix is to be had, I’m expecting a number of sites to pick this up quickly.
The Dash GPS takes this a lot further. It has Wi-Fi and GPRS built in, so you can send addresses from a Web page to the GPS - or find places in Yahoo! Local and navigate to them. The Dash unit has a model of traffic patterns in all major US cities built in, and when you ask for a route you’ll see several choices, with the predicted traffic on each one. As Dash users drive around the units send back anonymous information, so if several Dash users are stuck in traffic on one road you’ll get an updated traffic warning. If a new junction opens, the first Dash user to drive down it will add it to the map for every other user. If you don’t know where you’re going and you don’t know what the area you’re visiting looks like you can get a photo - from what’s tagged for that area on sites like flickr; if it’s not a helpful photo you can give it a low rating to help other users. 
If you search for a petrol station you get the price of petrol at each station listed. You can search for services and points of interest near your destination, so you can pick a parking spot or a coffee shop in advance. It’s less about just getting directions and much more about getting information you’d normally have to be at your computer to see.
As Dash’s Mark Williamson puts it “We’ve seen so much amazing information with geographical locatsions recently but the only way we’ve seen to access it is through the browser - what happens when you get in your car? Did you remember to print everything out before you left your PC? Dash allows you to take all of this amazing geolocated data off the browser and into the car. We think the future of content in the car is about letting you make decisions when you’re driving.” I hope that’s while you’re putting on your seatbelt rather than while you’re actually on the freeway.
So far there are 800 users between San Jose and San Francisco who’ve been driving around for eight months. Dash is recruiting another 2,000 testers and data collectors across the US with a plan for launching a US service later in the year - no dates for when they’ll come to the UK yet.
Dash completely anonymises the data that comes back from each GPS. John Krumm from Microsoft Research told the conference why that was a good thing. If you can get a trace of where my car has been at 3am every morning, you know where I live (or in our case where I’m staying this week). Drop that into a search engine and you can probably find my name, my shoe size and where my dog goes to school. Microsoft Research kitted out some volunteers with GPS trackers to see how well the directions you get from navigation software match the routes people drive when they’re not using GPS; answer, not very. “60% of the time the route people take is not the shortest or fastest,” says Krumm.
That might make for some changes in the next version of MapPoint and Streets and Trips, as might the technique they discovered for checking you’re on a road you could actually have reached in the time you’ve driven when GPS readings aren’t accurate enough to position you on the map. But while they had the data the team checked out how easy it would be to stalk someone from GPS readings, given that 219 people had offered to have two weeks of driving data collected in exchange for a one in a hundred chance of winning an MP3 player. The number of blocks of flats and underground parking garages in Seattle meant they only put names and addresses to 5% of the people they had traces for but in order to make sure that you can’t be identified, you’d have to add 5km of error to the GPS position or delete GPS information for 2km around your home address - which is a lot of errors and missing information.
And you will want services to know your location, whether that’s for improving the service you get by contributing information, insurance based on where you park and how far and fast you drive, traffic prediction, congestion pricing or the slew of location based services we’re going to see. Microsoft would like to be able to predict where you’re going to stop even if you haven’t punched in a destination. The annoying side of that will be adverts and coupons for local shops and services. But if you drive a hybrid car and you’re driving uphill - which would usually make the petrol engine switch on to give the battery enough charge to get to the top - it would be useful if the engine management system knew you were going to keep driving down the hill on the other side, where it could pick up some free electricity from regenerative braking.
What would you want your location to say about you?

One of the underlying themes at this year’s FiRe was the need to switch cars away from fossil fuels. What’s really surprising is that the work that’s being done to develop the batteries needed by tomorrow’s electric cars will also change the way we build and run our data centres.
Elon Musk’s Tesla electric sports car intends to change the way we think about electric cars. There’s a lot of work going into developing the motors and the control systems needed to build a car that out performs many of the petrol engined performance cars we see on Top Gear. It also looks good - the Tesla makes yesterday’s electric vehicles look like the milk floats they were (and does 0-60 in a handful seconds with a range of over 200 miles).
Tesla weren’t the only electric vehicle developer at FiRe. Battery manufacturer Altairnano were deminstrating a set of SUVs that here running off their new battery packs. As the batteries used new nanotech cathodes, they could be fully charged in less than 10 minutes - though not off your usual mains plug!
In one FiRe panel on the future of the car, attendees suggested that the future wasn’t just battery powered cars, but actually more hybridisation - cars would have not one, but two or three different power systems connected to one electric drive system You’d use the one that worked best for the task you had in mind - battery electric for in town, fuel cell electric for commuting, and petrol electric for long journeys.
So how does this all relate to the data centre? The key is the hybrid approach to power systems.
We run data centres quite inefficiently - pulling power off the mains when we need it, and delivering it to inefficient power supplies and voltage regulators. A report from Google and Intel shows that we lose nearly 80% of the power in data centres as heat due to conversion inefficiencies. There’s a lot of work currently taking place that suggests that running data centres off 48V DC power rails will offer the greatest power savings (and the least heat to be dealt with by even more inefficient air conditioning systems).
Altairnano are already delivering multi-kilowatt hour battery packs, and plan to deliver a gigawatt hour pack sometime in the next year. That’s not your standard battery, of course. It’s the sort of thing that will fill a room… But it’s also the sort of thing that can be used to power a data centre for many hours. These battery packs aren’t just generator replacements - you could also use them to moderate the flow of power into your data centre, charging the packs from the cheapest possible source when not much power is being used, and feeding it to your servers when its most needed.
We often think of green technologies as something that cost money - perhaps its time to start realising that the hybrid approach to delivering power isn’t only greener, when it comes to the data centre it can also save money!

We’re at the Future in Review conference in San Diego this week, listening to presentations and conversations on everything from the state of China and Japan to the nanophosphate cathode in the battery Altairnano is putting into new electric cars to Intel’s attempts to provide the healthcare equivalent of the PC - where the hospital is the equivalent of the mainframe - and that was just before the first coffee break.

Future in Review looks at the things that are happening now in all areas of technology that will make big changes in three to five years time. Ty Carlson, Microsoft’s director of technology strategy, talked about heterogeneous multicore PCs with eight and more processors. “We’re going to be bringing supercomputing to the PC and the mobile devices; we’re going to have amazing power on the desktop and mobile devices and in the megadatacentre. There’s been a lot of incremental updates to applications; we’re now faced with a sea change, we’re going so fast in this race that the applications have to be fundamentally redesigned.”

These days ‘sea change’ isn’t just a metaphor. Sequencing the yeast genome took a thousand scientists ten years and today we could do it in a single day. The reason the human genome could be sequenced so quickly wasn’t just faster computers; it was the algorithm of breaking the DNA up into fragments and sequencing enough of them to be statistically certain that we’d sequenced the whole genome and then putting in back in the right order.
Craig Ventner’s latest project is sailing around the world sampling barrels of sea water; by the time you filter out everything larger each barrel has around ten million viruses and from the barrels taken between the Galapagos and Halifax sequencing has turned up 6 million new genes. There’s a significant difference in the DNA found in different waters; 85% of the DNA is unique every 250 miles.
That’s a lot of new information for biologists and chemists looking for genes that could be used for drug treatments or even inserted into bacteria to replace industrial processes like fermenting alcohol to make fuel ethanol or converting the carbon in CO2 to protein or sugar. Ventner says “my goal is to replace the petrochemical industry by basically the next decade”. But the database is publicly available as part of the CAMERA (Community Cyberinfrastructure for Advanced Marine Microbial Research and Analysis) project at http://camera.calit2.net/index.php so students and researchers can work with the data.
The dataset doesn’t take up a lot of disk space – one satellite image of where a barrel of water came from occupies more storage than the entire database. But it took 100 million cpu hours to compute and several racks of servers are slicing and dicing the information so you can see how little generic difference there is between viruses that live in the balmy waters a barrel depth under the surface and organisms that live in extreme temperatures.
Ventner wants to move on from the “digitisation of biology” to synthetic genomics but Janine Benyus of the Biomimicry institute thinks there’s more digitising to be done. Biomimicry is innovation inspired by nature; she calls it “inviting nature to the design table”.
We could make buildings that wash themselves clean in the rain by imitating the surface lotus leaves. A firm in New Mexico is turning CO2 and sea water into limestone (instead of releasing 8% of the CO2 in the atmosphere by grinding up limestone to make cement), based on the way corals and shells form in sea water. The aptly named Frank Fish of Applied Fluids Engineering found than an airplane wing has 36% less friction and 6% more lift if you make it scalloped like a whale flipper rather than smooth. So far we can only make chips that put themselves together out of the silica in sea water in the lab but IBM has made a chip that runs 35% faster with less heat to dissipate because it self-assembles at the nanostructure.
There are plenty more good ideas in nature, waiting to be borrowed. It’s not even a matter of having the research,” says Benyus; “there are tons of biology papers sitting gathering dust– it’s literally an information problem”. The Biomimicry institute plans to make them available to engineers and materials scientists, but that means tagging them not just with the categories of biology they cover but with what engineering concepts they apply to. That’s like combining the categories on eBay with all the tags on flickr – and making them all make sense. The sea of information just keeps getting deeper.

WinHEC is the Windows Hardware Engineering Conference. It’s not the OS that makes the difference between the Mac and the PC - it’s the fact that you have a choice of hardware to build or buy a PC, and the fact that Microsoft has to support it all. Windows Vista has plenty of new features but many of them - from BitLocker to the Aero UI - rely on hardware.  This conference has been Microsoft’s first chance since the launch of Vista to thank the hardware partners who’ve come up with drivers - and to shake a finger at the ones who haven’t.
First Bill Gates and then Mike Nash both started by looking back at the launch of Windows Vista. Rehashing the marketing effort they put into Vista reminds hardware partners they can get some free marketing by labelling an application as Vista ready. Reminders of key technology that’s been covered at WinHEC before that’s become a significant part of the market for hardware partners says ‘trust us, we know what people want’. Listing the number of device drivers and certified apps doesn’t just give users some good news, it reminds developers that their competition is getting an advantage they’re not. And by saying, as Mike Nash did, that “This really is the year of focussing on the Windows Vista experience”, Microsoft is saying that Vista is what they’re going to carry on marketing this year and it’s time to get with the program and jump on the bandwagon - or get run over.
The other big themes of WinHEC have been Windows Server 2008 - now that beta 3 is out and the oh-so-surprising name is no longer a badly kept secret - and the 64-bit future. Windows Server 2008 has a raft of great features, from GUI-free Server Core to services loading in random addresses to improve security, and it’s Microsoft’s best story. There’s huge growth - the predictions are 8 million new Windows Server 2008 server boxes sold in 2010, far outstripping the 2 million Linux boxes predicted to sell in the same year. Apart from the virtualisation features that have fallen by the wayside -  which we’ll miss but then they were going to come for free - it’s on time, working well and making everyone happy.
It’s also the last 32-bit server operating system Microsoft will ever make. Windows Server 2008 R2 in 2009 will be 64-bit only.  If we don’t have more 64-bit drivers by then, those 2010 figures won’t include many copies of R2.
And if you take Bill Laing’s  phrasing in his keynote to be literally true - “this is the last 32-bit operating system we will produce” - then the next version of Vista after SP1 will be 64-bit only as well; after all, if the server team isn’t working on a 32-bit code tree, it would be a lot of work for the desktop team to keep working on 32-bit. Bill Gates has been saying for two years that it’s 64-bit time. The hardware is ready - but in many cases the drivers aren’t, so every keynote and presentation had a reminder, subtle or blatant, that it’s time to get coding.
Microsoft has been looking back to the Vista launch to remind hardware companies that Vista isn’t the future any more - 64 bit is.

SF writer Charlie Stross recently blogged a talk he gave to a German technology consultancy. It’s that rare and brave thing - an attempt at (not) predicting a future. In it Charlie pulls together a selection of trends, and suggests that the end point is a world where everything we do is recorded - so nothing is lost, nothing is forgotten.
Lifeblogging is already here. Nokia has basic lifelog software built into some of its phones, while GPS hardware is starting to appear in digital cameras. Meanwhile services like Twitter let us microblog our every move, and Upcoming lets people see just where we’re going to be. We might not have realised it, but we’re already giving away more information about ourselves than the governmant wants to store in ID cards.
If you read my personal blog you’d learn many of the things I like, and some that I dislike. You’d find out what I think about a mix of issues, see my photographs of places I’ve been, and read about problems I’ve had while writing programming tutorials. I’m sharing it all with hundreds, perhaps thousands, of people I don’t know, in parts of the world I’m unlikely to ever go. If someone else was collecting it, I’d be worried - but I’m giving it away of my own free will. You’d just have to visit a handful of other social networking sites to see my CV, see some details of my friends and colleagues - in fact to get as complete a picture of my life as a government security vetting.
We’re on the cusp of a major change in the definition of privacy. The models that No2ID and Privacy International have are being erased by millions of MySpace lifeloggers. It’s becoming trivially simple to trawl sites and use simple rules engines to infer information that’s not being given out.
Are Apples bloggers all working late? Perhaps something big is about to ship. Why have all the Microsoft attendees dropped out of one conference? Perhaps there’s an acquisition on the horizon. No one has given away anything specific - just a trained eye has put two and two together. It’s a process that can even be automated, looking for trends and reporting them…
Mary and I will be attending (and blogging) a futurist conference next week. It’ll be interesting to see what people there (including the author of The Transparent Society, an influential book on privacy) think about Charlie’s thoughts….
–Simon

One day your business is chugging along at the front of the pack, making money, winning awards and following the rules of the Harvard Business School by designing the best products you can for your best customers. Fast forward two years and you’re limping along at the back of the pack as some upstart takes the market in a completely different direction. You didn’t pay any attention to them when they started up because their products were so plainly inferior and your best customers assured you they wouldn’t touch them with a bargepole because they didn’t do what they needed; the upstart disdained your carefully crafted, proprietary, expensive or unavailable to licence architecture and threw together something modular that anyone could put together but no-one could make efficient because the components and the system were so crude.
But somewhere along the line people who would never have bought your products bought their new design even though it wasn’t at all powerful and it didn’t do most of what they wanted either, but it did something and it was cheap enough to put on expenses. And every few months, while you honed your team and polished your master plan and prepared for the next three year design cycle, the new guys put out a newer version - and each time it did a little bit more. Eventually they had something good enough for your customers too, at a fraction of the price you charged, so you looked into making a product to compete directly. But your profit margins at the top end of the market were 60% and theirs were 40% or even 20% and you didn’t see the point of undercutting yourself to make lower margins on cheaper products and defend the least profitable part of your businesses. And then most of your customer stopped returning your calls and professors at the Harvard Business School started asking how a smart company like you got so stupid so quickly.
That’s what Clayton Christensen (one of those professors at the Harvard Business School) believes happened to Digital when the PC came along and to Dell when notebooks took over from desktop PCs. The only reason it didn’t happen to IBM when the PC came along was because IBM created new business units with the mandate to make money at the lower margin - and to kill the parent company and take over from it. The sustained innovations that make better and better products to sell for higher and higher prices to existing customers don’t stand up against disruptive innovations that start by being too basic for even the least advanced users but improve quickly without most people noticing and win over people who weren’t customers of any of the competition because even the least advanced of the old products were beyond them.
When that happens, the market changes under you and the rules change too. With a proprietary architecture like IBM mainframes or Apple Macs, the companies that make components don’t make much money because the value is in how IBM and Apple put the systems together. With the PC, the manufacturers of the CPU and the graphics card and the memory make more of the profits, and anyone can put a PC together. IBM had better operating system technology than Microsoft and better microprocessor technology than Intel - but it chose to outsource those to concentrate on design and assembly because that’s where the company had made money before - so IBM managed to hang on to the part of the system that wasn’t going to make money.
At this point Wall Street starts pushing companies to get their costs down or get rid of some of those expensive manufacturing assets where they make the PCs. Christensen tells a nice story of how Flextronics hollowed out Compaq by offering them better and better deals to take over first supplying the motherboard, then assembling the system and then supply chain management and finally design. Each time Compaq’s revenue stayed the same but profits went up because of shifting the assets for each part of the process off the balance sheet, along with the 20% discount Flextronics gave them each time and the shares did well “especially as Wall Street loves asset light companies”, while revenue and profit went up for Flextronics, along with their share price “because Wall Street loves value added companies”. By the end of it Compaq had nothing but the brand of their PCs - at which point Flextronics went to Best Buy and offered them a 20% discount to take the PCs they were building direct and cut out Compaq altogether…
Which brings us to Apple - and RIM, as Christensen was telling these stories at RIM’s Wireless Enterprise Symposium in Orlando. Allowing for the fact that keynote speakers don’t usually turn on their hosts, the professor was still dismissive of Apple’s chances because he believes they’re going for the niche end of the market. “The iPhone is really a sustaining innovation relative to the business models of Nokia, Motorola and Samsung. It might be just the most miraculously designed product ever, but they’ve picked off a piece of real estate at the high end where Nokia, Motorola and Samsung can fight back and the fact that they’re motivated to go after that is clear.”
And while Apple is busy commemorating the sale of the 100 millionth iPod, Christensen thinks the recent drop in market share in favour of SanDisk and Creative isn’t what Steve Jobs should be worrying about. “Early MP3 players were clunky and hard to use; Apple brought in a proprietary architecture and did very well. But if you look at the adverts in the paper today, almost every phone proclaims that it plays MP3s. Telephones as music players don’t perform nearly as well as a proprietary architecture but in a few years time you can expect that they will. If I were running Apple, as I see these disruptive phones coming in I’d say the money migrates to the one who provides the enabling subsystem and I’d want to make darn sure that all of these phones were downloading their tracks from iTunes.”
And what about the proprietary BlackBerry architecture? “RIM does the same thing but in order to handle the always-on facility the architecture is excruciatingly interdependent - in order to handle primarily the power management problem.” As in, you can have push email and four day battery life today as long as you have a BlackBerry and not a Windows Mobile phone. And tomorrow? “What RIM needs to worry about at what point does a non-integrated modular architecture make sense.” As in, when networks and processors get faster and use less power, Windows Mobile or Linux could do the job more cheaply if not as well.
And given that the company invited Christensen to talk about this and must have come across this theories of disruptive innovation before, have they learned from history? “RIM is making their operating system more available to other manufacturers. The optimised architecture of the BlackBerry is a temporary advantage and they want to make sure that whoever is successful uses the secret sauce of the RIM operating system.” It might come from Motorola or HTC, it might run Linux or Symbian or whatever Palm OS mutates into, but if it runs the BlackBerry software as well, RIM will carry on being the byword for being connected.
So, do you think RIM will live happily ever after?

RFID chips let businesses track products and shipments accurately. If you’re buying rock from Graniterock your truck gets tagged by RFID which tells the weighbridge what kind of marble you have so it can work out the price with the weight and send the invoice as your delivery is on its way to you. Luggage going round Hong Kong airport has an RFID tag so it’s harder to lose. BP tags canisters for liquid gas so they know when they’ve been stored safely - if you can’t prove the canister hasn’t been exposed to unsafe conditions you have to spend the money testing something that’s probably OK before you can use it. That’s RFID making things better for business and maybe saving enough money to keep prices down.

But the idea of embedding an RFID chip in DVDs, MP3 players, ink-jet cartridges, flash memory, electric toothbrushes and all those other consumer goods that are small enough to steal isn’t quite the same thing. NXP - a spin-out of Philips - wants to sell more RFID tags; Kestrel wants more people to use its Radio Frequency Activation technology which will turn the gadget you buy from a brick to a useful tool only when you pay for it. Retailers and manufacturers want to avoid shop-lifting and boxes that fall off the back of the lorry and end up in your local market. And while I’m all in favour of removing security devices and those impossible to open sealed plastic packages in favour of more environmentally friendly packaging - or frankly, just less packaging filling up my rubbish bin - and I’m sure that the manufacturers up prices to cover what might get stolen, I don’t think this is the way to do it.

According to the press release, Christophe Duverne at NXP says “RFID [makes] the shopping experience more enjoyable for consumers”. Does that include the experience when the RFID chip which you have just made the single point of failure in my new purchase malfunctions and my gadget turns back into a brick? Consumer electronics have a short enough life span that adding an extra component with no function other than to stop things working strikes me as a terrible idea. If your new gadget breaks sooner, you’re just putting the gadget into the landfill instead of the packaging. And that’s just things failing and breaking; what about the potential for denial of service attacks and subscription pricing that holds users to ransom?

Would you rather pay a little less for a DVD or an ink cartridge that has to be turned on before you’re allowed to use what you’ve already paid for?

-Mary

While we’re on the road, we’ll only be logging in remotely and we won’t be at our desks. Wouldn’t it be nice if the server knew that automatically and blocked local logins until we return? It’s the kind of simple, common sense measure that IT isn’t usually good at but that can make a big difference to security. If we used security fobs to get into the office, we could use Imprivata’s OneSign to tie login access to physical access, so we couldn’t sign in from a desktop if we hadn’t already fobbed into the office and we couldn’t sign in on a remote connection if we had.
 
OneSign started as a single sign-on appliance. By simplifying the tools for managing single sign-on, application users are linked to one specific log-on account. It’s easier to manage access – and if someone leaves the company, shutting down their network password means they can’t use any applications and walk away with business-critical data. You don’t even need a password to use OneSign – you can use it with token-based authentication systems or with smart cards.

Many large businesses use keycards to control access to office space. Cards can be used to limit access to specific areas (ideal if you need to keep traders and analysts apart). Simple readers control door locks and pass information onto a central server. You can be sure that someone is in the building (or at least that someone is using their card) if they’ve carded in through a door. Bringing tools like OneSign and door readers together is an interesting move – it allows system administrators to lock applications down by location as well as by user.

OneSign now uses door system databases to indicate that a user is in the building, and card readers attached to PCs to authenticate access with a swipe of the same card. Application access will only be authorised if a user is in the building. If a card hasn’t been swiped at the door, there’s no way in to the code. One sort of access control adds extra features to another…

This fits in very well with the idea of contextual security. If I’m at my desk in my office, I should have higher access rights than if I’m using my laptop and a VPN over coffee shop WiFi. I should have even less access if I’m at an internet cafe, and fewer still if I’m in some part of the world with a poor reputation for internet security. Context is hard to define, but OneSign’s integration with building access control systems goes a long way to solving part of the problem.

–Simon

We’re off on our travels again this week, heading to the US for a set of conferences and meetings. That means jet lag, long days and dealing with time zones. One BlackBerry feature I’ve been missing on the HTC Excalibur I’ve been using lately is the ability to turn the device on and off automatically; that means I can avoid 3am phone calls from UK colleagues who forget I’m in the US but not miss the alarm I’ve set for the morning or have to wait till I get up to download email.
I’m trying out a couple of Windows Mobile smartphone utilities from Connective Tools to get the same effect. SleepWell! can turn my smartphone off automatically at a set time, or switch it in and out of Flight Mode - and turn Bluetooth on or off as well; that’s more flexible than the BlackBerry that can only turn on or off but not as powerful because SleepWell! can turn my phone off but not on again. So I’m switching it into Flight Mode at 3am with SleepWell! and using another tool called SwitchProfile! to switch the sound profile  to Silent mode at 8.27am then SleepWell! turns Flight Mode off at 8.28am - email starts arriving but I won’t hear any phone calls until I’m ready to talk to people at 9.30am when SwitchProfile! puts  it back to full volume. I’m also using AlarmMaster to remind of things I don’t want to put in my calendar, like when office hours start in the UK so I don’t wake anyone up back home.
They’re very simple utilities - once you realise you use the Weekly option to set daily repeats, which makes it very easy to make weekday and weekend times different and get a lie in - but also very useful. SleepWell! is free and the others are only a few dollars each ($3.25 to $8.95). And not getting woken up the first time you finally manage to sleep through the night when you’re travelling? Priceless.