HOW AGILE IS YOUR TEAM? I am often asked the above question. The answer for my innovation team is not much. This is because we develop hardware-intensive technologies, and apart from a few exceptions, it is not possible to adapt *continuously* the technology to *end user* feedback. Instead, we cherish and process the feedback and iterate as fast as permitted with hardware (e.g. 1-2 months for PCB design or chip assembly). In addition, *end-user* feedback is often hard to collect for semiconductor components as chips are hidden deep inside the tech stack. Instead, we collect feedback from business stakeholders along the value chain -- which is still valuable. Finally, we don't ship incremental product to end users. Instead, we develop lab prototypes to show the technology to early adopters. In summary, with respect to the ideal agile development, our development approach exhibits the following deviations: - iterative (cycle time ~ 2 months) instead of continuous value delivery - feedback from business stakeholders instead of end-users - lab prototypes of increasing fidelity instead of operational product increments All in all, this represents a decent compromise between the rigidity of plan-driven development and the ideal agile development (which is hardly applicable as-is in my humble experience). It provides a decent amount of risk reduction through the learning cycles. The risk of product misfit remains as the end-user feedback is relatively weak. Please challenge while remaining respectful. #innovation #hardware #agile #semiconductor #design #team
Gaël Close’s Post
More Relevant Posts
-
I’m #hiring a biosensor engineer for the MELEXIS Innovation Lab. We look for individuals with a strong track record in applied research, with hands-on experience at the intersection of biochemistry and electronics hardware. An entrepreneurial mindset is also desirable to mature the technology and bring it to market and make a real-world impact. The position will offer plenty of learning opportunities: at the bench with state-of-the-art semiconductor chips functionalized to target specific biomarkers, but also in the commercial aspects. The position is open at multiple ranks. Fresh graduates from a biomedical engineering degree, as well as seasoned professionals, are encouraged to apply and anything in between. The role will be adapted accordingly. The position is located in the MELEXIS Innovation Lab, in Bevaix Switzerland. Lake & mountain views are included. The Lab was established two years ago and focuses on advanced R&D in robotics, alternative mobility, and digital health. The Lab conducts research projects together with academic and industrial partners, with a vision to incubate internal start-ups. It is an excellent place to launch an industrial career in applied research.
This content isn’t available here
Access this content and more in the LinkedIn app
To view or add a comment, sign in
-
People sometimes ask me why Switzerland is an innovative country... Today, I had the pleasure of participating in an open innovation workshop about MEMS sensors for bioreactors. I learned that bioreactors are producing drugs, vaccines, food, ... They are also sensor-rich applications. It is also a sizeable business (>10 Bn €) and growing fast. This is a fertile ground for sensor innovation, with clear links towards established Swiss strengths (biotech, microsystems, etc.). The event regrouped nine participants (academics, startups, SME, industry, ...) covering the value chain from sensor providers (MELEXIS, Innovative Sensor Technology IST AG) to bioreactor manufacturers, with skilled microtechnologists from the watch industry. It was supported by Innosuisse and their Microtech booster program. It took place in Zurich, an proven innovation hub easily accessible by train. With minimum formality, Innosuisse provided seed funding for a pre-feasibility study & prototyping. Two innovation coaches were also available to help the participants reach a consensus on the innovation problem at hand. Together, we brainstormed using low-tech tools: Lego and even Duplo using the SeriousPlay method. This method was pioneered by Lego and IMD in Lausanne. The participants sketched possible solutions with Lego blocks. Suddenly, the barriers between disciplines broke down. Highly technical concepts like field-effect biosensors, MEMS, microfluidic channels, pumps, metabolism, cell culture, autoclaves, and in-line bioprocess analysis came to life in the forms of Lego blocks. It was an eye-opener, aligning stakeholders with different perspectives. We consolidated the agreements into a graphical innovation canvas. The next step is to actually BUILD the most promising idea, and learn together from the joint experiment. We might then converge towards a focus high-tech R&D project, pushing the state of the art with clear business impact. So why is Switzerland innovative? #seedfunding #SME #hightech #lowtech #coach #biotech #sensors #MEMS #hub #science #business #watch #microtech
You recognize what these legos represent, don't you? You can of course see the bioreactor, the sensing module with fluidics, and of course the customer who's very happy with a great added value. The elephant drives a mass-manufactory line. This was the work at the kick-off of the project team from Samuel Huber and Yves Mermoud, selected for the 6th Microtech Booster loop, with the support of Innosuisse. Today's workshop enabled the team to: -Get to know each other and have FUN! -Understand each other's expectations and competencies with the legos serious game, -Explore the challenge to be worked on, and finally define an "Innovation Project Canvas". The team, made up of 10 entities has invested in total 11'000CHF in this project, to which the Microtech Booster is adding 20'000CHF for their 8-month feasibility study. The Microtech Innovation Booster gives them the opportunity to: -Participate in a collaborative and innovative feasibility study, with reduced risks, costs, and shared resources, -Share competencies with other academic and industrial partners, -Enlarge their network, -Discover innovation tools. Congratulations also to the institute IDEE from the OST – Ostschweizer Fachhochschule for this great managed kick-off meeting, using the Legos serious game. We look forward to seeing the results of their study at the end of the summer! Best wishes to the whole team! Christoph Hepp, Davide Migliorelli, Caspar Demuth, Dirk Hebel, Julia Schläpfer, Léonard Barras, Juan Gualberto Limon Petersen, Iris Poggendorf, Nicolas Richerdt, Sébastien Walpen, Thomas Valentin, Yves Pelletier, Andreas Bauer, Axel Fanget, Barbara Horvath, Gaël Close, Samuel Böhni, Lukas Schmid, Loïc .
-
+5
To view or add a comment, sign in
-
Do you want your next paper and/or poster to be fully #reproducible from #Python code & data? I created the following skeleton on Github to save time when preparing a reproducible computational paper. The skeleton sets the structure for modular Python development (reusable modules, unit tests, notebooks). The companion paper and poster are drafted in easy-to-learn #Markdown. The skeleton contains the commands to generate the paper and poster via #Quarto in both HTML and PDF (with simple commands, e.g. `run pdf`, `run poster`). The paper is formated in #IEEE style, and many more journal formats are supported in Quarto. https://lnkd.in/dSdzcESV
To view or add a comment, sign in
-
The MELEXIS Innovation Lab just demonstrated at the IEEE SENSORS conference a novel encoder with 17bit of RMS angular resolution (in a 500-Hz bandwidth). This encoder is well suited for precise robotic applications. For example, the measurement of a joint position on the gearbox side. After a few years of incubation, this innovation is ready to leave the lab. We will soon release an evaluation kit for early adopters. Get in touch if your next project needs a precise & compact encoder at a competitive price point. Let's democtatize robotics together with novel magnetic sensors.
Last week on the IEEE Sensors 2023 in Vienna/Austria Bruno Brajon and I presented our "Novel Rotary Encoder with Multi-Axis Hall Sensors", a project from the #MelexisInnovationLab, in a Live Demo Session. During more than two hours we explained and discussed with students, professors and professionals - what an experience! In then end we were additionally honored with our selection on the best awards finalists shortlist. Thanks a lot to Gaël Close, Eugene Lomakin and all our Melexis colleagues who contributed to get the this thrilling demo running.
To view or add a comment, sign in
-
The #Melexis Innovation Lab has just published today a survey paper on integrated magnetic sensors. Such sensors are ubiquitous, with 10 billion sensors deployed annually. The paper provides a concise overview of the state of the art. It covers multiple technologies applicable to a broad range of magnetic fields and field gradients. This is relevant for applications such as position sensing in cars and robots, current sensing in electrical vehicles, and biomagnetic sensing. The paper is published in the journal IEEE Access: * Bruno Brajon and Enrico Gasparin and G. Close, “A Benchmark of Integrated Magnetometers and Magnetic Gradiometers,” IEEE Access, vol. 11, pp. 115635–115643, 2023. Available: https://lnkd.in/eUvXkCgH The paper is accompanied by a computational Python notebook hosted on Code Ocean, allowing the community to reproduce and even extend our analysis by incorporating other devices. We intend to maintain the benchmark by updating the computation notebook as new devices emerge (e.g. graphene Hall sensors, …). Notes. This work was inspired by the "ADC Performance Survey 1997-2023" by Boris Murmann (https://lnkd.in/etc4GTWr), and the "Smart Temperature Sensor Survey" by Kofi Makinwa (https://lnkd.in/epnrvDrb). Given that magnetic sensors are implemented and reported in vastly different form (full smart sensor versus bare transducer), we didn’t compare the energy per conversion step, unlike in these two references. Instead, we focused on the (simpler) trade-off precision vs range—ignoring energy consideration for now. There is room for improvement and future work.
A Benchmark of Integrated Magnetometers and Magnetic Gradiometers
ieeexplore.ieee.org
To view or add a comment, sign in
-
Just finished a 3-day course at Formation Continue UNIL-EPFL on AI and the Internet of Things. A mix of theory with hands-on exercises. As a sensor developer, this was good to see how the sensor data are communicated, aggregated on the could and exploited. As a manager, it was also good to go through the process hands-on. For the IoT part, we spent some time playing with a temperature sensor, communicating wirelessly using the LoRa protocol (a good fit for low-bandwidth applications, like asset tracking) to a nearby gateway. The data payload was aggregated and stored on a server, then available via a web browser. For the AI part, we had a glimpse of some hardware accelerator chip, designed in Europe, made in Europe based on a RISC-5 platform. An engineering marvel capable of running AI algorithm right at the edge, albeit more challenging to operate. We also went through the online tutorials offered by Google (Teachable Machine) and Amazon which have lowered the barrier considerably to get neural network running (for demos). I am looking forward to sharing my takeaways with my team at #Melexis. We have the good habit of always sharing the key learnings in a "Tech Talk" within the team anytime a team member goes to a conference or short course. Thanks to the instructors David Atienza and Andreas Burg and the staff.
To view or add a comment, sign in
Enterprise Business Architect
1yHi Gaël, nice work! The question is: what is "ideal"? I guess context is crucial. To me, it is really a matter of mindset. Trying things and learning from that, to gradually move to a better (more agile) way of working. Good luck with the journey ;-)